Some Number Crunching / understanding tourism output and GM strength calculation.
There are 3 layers to tourism that I will respectively name raw, base and civ total.
The first layer, "raw" includes everything generates +TPT. This can be further broken down into great works, theming bonuses, tile improvements(incl landmarks)/wonders/natural wonders bonuses.
The second layer, "base" adds the multiplier bonuses that are inherent to your own cities or your own empire. GM tourism strength is calculated from this "base" number.
The third layer, "civ total", adds the multiplier bonuses to each civ. This includes OB, Diplomats, Shared religion, Shared ideology, Trade routes.
First Layer
Each GW is worth 2. It is improved to 3 with Hotels, to 4 with Airports and to 6 TPT in capital from NVC (national visitor center). This means that if, somehow, you went deep Piety and took the "can buy any GP with faith" reformation bonus and eventually ended up not completely filling the aesthetics tree, putting GW in your capital at the cost of other cities' theming bonuses will grant more TPT. Particularly more if you have enacted your religion. In general, at least on deity, you won't get to hard build wonders with GW slots in your satellite cities and if you have a GE, it's obviously best to build it in the capital.
Theming bonuses are worth 1 TPT per GW slot. There is some twist where the louvre can achieve a partial bonus and only provide +2. I have not seen it on other wonders. This bumps up to *2 with either completing aesthetic tree or playing as France or *3 with both. (if anyone has a late game save as napoleon, I actually need to verify that the bonuses are additive as with most other % based bonuses in the game)
Landmarks/World Wonders/Tile improvements/Natural Wonders
Each culture point provided by any of the above is worth 0 until hotels to which it shifts to .5 and up to 1 TPT per CPT with airports. As with GW scaling, the NVC improves this to 2 TPT per CPT. This give a strong edge to France that can simply spam chateaus on all tiles just a few turns before the Internet to offset badluck on wonders are to improve the odds of catching up a runaway with just the GMs coming shortly after.
Second Layer
City bonuses or empire bonuses.
The Internet empire wide +100%.
WC religion Capital bonus +50%
T3 Freedom tenet broadcast towers +34% in each city
International game first place +100% for 20 turns
Brazil's UA +100% for the golden age duration
All of these bonuses, even the empire-wide bonuses like The Internet are calculated at the city-level and are
additive. All these bonuses are applied before GM calculation. The number at the top of the interface includes any such modifier. Which is the same as saying that your GM power will be 10* whatever is listed at the top of your screen. Not that end-of-turns occur in sequence and that Great Persons are calculated and born BEFORE science. Thus, if you time a GM to pop the same turn The Internet is teched, you won't receive the tech bonus on that GM. It has to pop the turn after.
Third Layer
Targeted civ dependant bonuses. These are to be seen in the culture overview tab - influence by player subtab.
+25% bonus for Open Borders
+25% bonus for Shared Religion*
+34% For shared Ideology/ -34% for different ideology
+25% for trade routes between the civs
+25% propaganda from diplomats
+50% tourism when at war with a common for (Autocracy T3 tenet)
+34% tourism to other order civs (order tenet - it basically double dibs the shared ideology bonus)
+34% tourism to civs that have less happiness (order tenet)
Do not get duped in thinking that autocracy/order provide more tourism than freedom as not only is the +34% from broadcast towers guaranteed as it is within your empire, but it also enhances your GM power as it is in the second layer of tourism calculation.
*The 25% bonus for shared religion disappears once a religion has been enacted as World Religion. However, this does not mean that enacting is weak. For this 25% bonus to apply, the 2 civs capitals need to share the same religion (as a dominant religion). Thus, on a typical 8 civ map with 5 religion, it means that you will really only get the 25% bonus on at best 3 other civs whereas the enacted 50% applies abroad to all 7 enemy civs.
Each of these bonuses are additive with each other but are multiplicative with the base tourism (thus multiplicative with second layer %scales). However, they have no impact whatsoever on the power of GMs.
Discussion on the numbers
What does the above numbers mean strategy wise?
It means that strictly from the point of view of maxing TPT from GMs immediately after The Internet, freedom is far superior because both the +4 from landmarks and the 34% from broadcast tower apply to raw and base tourism and thus boost GM output. However, from a turn-to-victory min/maxing perspective, Order will typically provide an earlier turn-to the internet and thus better turn-to-victory. It requires the player to have basically been wonder spamming though to offset the loss of GM strength and thus is not recommended as the generic strategy in the guide.
It means that enacting a religion can significantly offset having had a game in terms of wonders built. (284%/234%-100%) = 21.4%. Let's numb it down to 20% or so since it only applies to the capital raw tourism, this means that enacting your religion will increase your late game tourism/GM strength by 20%. Sadly, if +3 culture per wonder was enacted super early, it make it extremely hard for a player with very few wonders to eventually catch up as then its late game TPT is capped much lower than if he had spammed wonders, even if he manages to enact his religion.
It also helps answering questions like what wonders are essential and can I pull this off if I missed a big one. Let's take what I consider to be "the basic setup". That is, the player builds Sistine, Uffizi, Louvre in the capital and no other wonder. Runs 3 city. Does not have access to cathedrals but fills out all of the GW/GA or artifact slots and does not produce a single GM in the early game to save them for late game runaways.
Worst case wonder scenario TPT - NO THEMING BONUSES.
Assume roughly 3 landmarks none of which are within capital 3 tiles range (so they only get +100% bonus and not NVC)
3 cities
all museum filled with artifacts
Essential wonders
Sistine 2 GW +1 base cpt
Uffizi 3 GW +2 base cpt
Louvre 4 GW +2 base cpt
museum 2 GW
Amphitheater 1 GW (w)
Palace 1 GW
Oxford 2 GW (w)
Nat Epic 1 GW (w)
Heroic epic 1 GW (w)
Hermitage 3 GW
3 cities total
capital has 20 GW for 120 total base +28 base from wonder culture (with +3 culture per wonder enacted)
satellite cities together have
6 gw for 24 total + 30 base from landmarks
Grand total raw
178
after 134% bonus (no enacted religion)
416 TPT (base).
This is fairly low. Typically, in fact especially if the +3 culture per wonder or +2 culture per settled GP/+4 per landmark have been enacted early (effectively creating some runaways), this won't be enough to overcome the deficit to "culture runaways" with 4 GMs (one pop, 3 from faith) and thus will require some GnK style "spam next turn" as well as a few additional GMs from GPP to close the game. If you add "side wonders". Adding an enacted religion to the count bumps your base TPT to 505 which will be enough to finish most games where you reach the internet somewhere between T250-T265. If it goes on longer, since tourism must overcome game-long culture, you will either have to do some spam next turn or attempt to achieve slightly higher TPT.
With about 600 base TPT @ the internet, one can be confident that 4 GMs will be enough to close the game. The only exception is, again, if cultural heritage was enacted around T150 or before because then the wonderful civs become one or a few runaways. It is extremely hard to counter proposals in the first 2 WCs and as a honest statement about CV, if cultural heritage is enacted first WC and the human player does not manage to wonder spam, it would be wise to change VC objective. 600 TPT can be achieved with the above + some theming bonuses + maybe 1-2 more random wonders with some + culture. Fairly easy ones to get are: Oracle (unless Mayans are on the map), Porcelain Tower (with -33% to scientists/ge/gm enacted early, most civs tend not to take rationalism), Big Ben (it costs 1 SP but it tends to be leftover somewhat late so it can give an extra way out to the desperate human player. I don't recommend it however), Status of Liberty (very easy, rarely more than 1 competitor and typically its one of the later civs techwise) Eiffel Tower (this one tends to hurt diplomacy a little bit though). Cristo Redentor or Neuschwanstein.
Why is religion so important even when you don't have enough CS allies to enact it in the late game? Any faith beyond the 5k for 3 GM can be spent on GEs to catch either the few harder to get wonders (namely catching both Uffizi and the Louvre can be a pain without any GE. Especially if you have not yet made it to 1st or 2nd in tech) by architechture). As a matter of fact, If you delay all GMs until after the Internet, your 2nd GPP GM will kick in about 10 turns after the first and one could consider reallocating 2500 faith to purchase a 3rd GM into 2 GEs in the industrial/modern era. One for the louvre and another one, probably for Eiffel or Cristo Redentor. The raw tourism difference between Eiffel and cristo redentor is of only 4 because of the double dib TPT per CPT from NVC in the capital so SP timing might justify using a GE on CR over Eiffel.
An additional reason to get a religion is that it allows you to produce GPs and thus also to settle GPs. This means a few additional tiles possibly producing +2 culture (effectively +4 raw tourism in the capital) from that WC resolution. It also means that it is much easier to produce the necessary total faith between settled GPs, +4 faith per settled GP from freedom tenet and the +8 FPT provided by the Grand Temple. Last, it lets you pick capital-centric follower bonuses to help your few city-tall game (RC, swords into plowshare, Divine Inspiration, Religious Art). Finally, it gives you an additional way out if you failed miserably at wonders by filling out the piety tree for typically an additional +24 raw tourism (as much as what the Louvre produces pre theming bonus) from ~ 4 settled GP. It is not the recommended path because of the SP competition but can be a solution for slow games (games that would end beyond T300).
Is it still possible to produce 5k faith over ~100 turns or so for 3 faith GMs without a faith? Short answer is yes. Long answer is it requires you to get Patronage/Consulates early to get the faith from CSs to compound early. However, it more or less requires you to befriend these CSs and begin to pile FPT from shrine/temple as soon as the 5th religion was founded (as since you can't produce GPs randomly anymore, you can begin to pile up rather than only be able to pile up shortly before industrial era). It also means that you will need to put considerable CS effort into allying faith CSs as opposed to simply befriending them. Sadly, unlike going for 5th religion, it neither gives you a capital-centric religion nor nearly as much risk-control. Befriending/allying 4 faith CSs isn't too hard if there are 4 faith CSs on the map. However, if there are fewer or if Ghengis decided to raid them all, you are put in a situation where you might want to go for a different VC whereas catching 5th religion prevents that.
But then how can I achieve 5th religion? It seems so hard when I can't get a FPT pantheon.
Indeed it is but here is an opener that will allow you to get pantheon + 200 faith somewhere between T70-T75 in order to pop 4th or 5th religion before T80. It does NOT guarantee a religion. In some maps where 2 of (Ethiopia, Byzantium, Maya, Celts), the 5th religion will go before T70. I've spent all of last week iterating over this and I would say that I managed to catch 5th religion in roughly 75% of the games. Sadly, it slows the opener slightly more than a regular full trad opener and thus if you chose to take that path and miss 5th religion, you should discard CV and aim for either diplo or science from that point on...or simply reroll.
The opener
Steal 1-2 worker from preferably a nearby AI, worst case scenario from a CS. You will need to improve at least one luxury somewhere between T35-40.
BO - scout - monument (if culture ruin was poped by the warrior, do a 2nd scout) - shrine - granary - settler*
* if your capital has 4 +2h tiles that can be worked, shift to settler after shrine, as soon as cap is size 4+. However, since most capitals will be settled near mostly food tiles, completing the granary will often result in 1-2 more hammers for settler building. The objective is to attempt to get settler built in 8-10 turns. Note that if your capital has no 3f tile access at all and no wheat/deer to compound from the granary, it delays the timing just enough to reduce the success rate to 40-50% 
SP order
If you hard build the monument, SP order is trad opener -> piety opener -> Organized Religion -> Regular full trad
If you picked 20 culture ruin early, SP order is trad opener -> Legalism -> Piety opener -> Organized Religion -> Regular full trad
Tech order is
Pottery -> Lux techs -> beeline philosophy.
As soon as philosophy is completed, shift your build in capital and 2nd city to Temple, focusing production. If you had managed to get 500g and bought 3rd settler early enough to get a shrine up, you can skip building the temple in the 2nd city as it will generally finish building it somewhere between T65-T75. anyway.
I will eventually provide additional details with a game example as to what to build "between" first settler and temple but it is highly dependant on the map and rng. Rule of the thumb is library if I have no neighbor within 10 tiles of either my 2 cities, caravan otherwise to help speed up the first DoF, get more BPT than the library and some GPT probably to bribe a warmongerer neighbor against a different civ, a 3rd settler if I need to claim land asap and am already happiness capped to grow but expect to improve luxuries in the mean time, worker otherwise unless I somehow magically stole 4 workers already. Rarely, I will build an archer not so much for defense as this is what bribing is for but in order to catch a few CS quests. Particularly culture/faith CSs (culture to reduce the impact of delaying tradition completion by going piety).
