My point with pushing tech was simply that the sooner you have NVC and double tourism output, the sooner you start overtaking them. It makes more of a difference than any one other thing you can do because even IG isn't permanent. Yes, clearly you have to get great works or some other source of tourism, you clearly can't ignore it, but the sooner you hit The Internet, the better, unless you're playing as a civ like Brazil, France or Polynesia who can hit such high tourism levels pre-internet that they don't necessarily need it to win. Delaying the internet only makes it harder to catch up, and I just assumed we were talking about bulbing musicians post-internet, here. I mean, I take it as a given on deity that you would need to do that in most games to overtake any cultural powerhouse you weren't able to slow down in some way.
Similarly, of course I'm not suggesting that you don't need key wonders, or that you don't need to block cultural runaways somehow. But it's still all out tech to internet at the same time. Sure, maybe you don't hit internet as fast as humanly possible if you're building archaeologists and museums, but, you need to tech rush to even have a chance at digging enough artifacts in a deity game. And getting wonders is easier if you hit techs first... My point was that you can't, say, prioritize tourism over tech and have a good chance of winning on deity. In the old cv, you just had to get massive cpt, which didn't require getting much farther than the industrial era. As opposed to requiring near end-game tech ASAP... Clearly it's now an SV. But whatevs.
Don't get me wrong, I like the idea of tourism and the new CV in general, I just don't like that there's a win button at the end of the tech tree, or that tourism is so hard to get early that you have a massive hole to dig out of once you hit post-industrial. The amount of tpt you can generate pre-hotels and airports is trivial for most civs.
For CV you need BOTH, I think is what we're trying to say.
Blindly pushing tech isn't going to work either, but so is ignoring tech.
Let's use a simple example...
Against a (somewhat above-average) target civ which has 300 cpt and has a 25k headstart at t200. For the sake of simplicity let's assume this is a duel map and you don't have to worry about other civs.
A) You reach internet rather quickly at t200 ignoring everything else but science... your tpt is 400 (and I think I'm being quite reasonable with this number, if not a bit generous). Let's say you have +70% modifier (and here I'm assuming you got +15% from OB and routes from aesthetics);
You need to close the gap of 25k using your spillover of (400x1.7)-300 = 380.
25k/0.38k = 65 turns. So you finish at t265;
each GM of power 4000 shaves off an equivalent of 4000/380 = 10.5 turns. You spawn 2 GM, and you finish at t244 or so.
B) You reached internet at an above average time at t220, but your tpt is 600.
Spillover of 600x1.7 - 300 = 0.72k
The gap is closed within (25 + 20x0.3)/0.72 = 43 turns; you finish at t263,
with each GM of power 6000 shaving off an equivalent of 6000/720 = 8.33 turns. 2 GM and you finish at t245 or so.
C) You reached internet after going for Hubble and bulbing at t240, your tpt is now 800.
Spill over 800x1.7 - 300 = 1.06k
Gap is closed in (25+ 40x0.3)/1.06 = 35 turns, you finish at 275.
Each GM of power 8000 shaves off the equivalent of 7.5 turns. 2 GM and you finish at t253.
C) seems the worst however...
As you can see, the more tourism after +modifiers you have, the less the relative power of GMs (each GM shaves off less turns); GMs are worth more turns on civs who hate you basically. Conversely, if you've had a sub-par game where your max tourism is low, then faith becomes all the more important.
Also, the lower AI's culture output, the more effective fast teching becomes. Conversely if their culture output is high then the final maximum number of tourism is supreme. If we changed the math such that that same AI now has 500 instead of 300 cpt for example...
A) finishes at t338, each GM for 22.
B) finishes at t287, each GM for 11.5
C) finishes at t292, each GM for 9.3
Here suddenly A is the worst.
Keep in mind that I completely ignored the tourism B and C gets during those 20 and 40 turns before they reached internet respectively.
If cpt is 700 (say a complete runaway) only C has a reasonable chance of finishing CV.