In my opinion, the only answer is making small empires viable. I'd wish I could somehow blast a large-map game in high difficulties with a small empire, perhaps 2 city 2 puppets. But that would be no doubt hard to implement properly... so I guess we're all stuck.
It's my proposed solution in the Ideas & Suggestions forum, with a post that is fairly longer than this OP.
There's a lot of mechanics that prevent a true exponential snoball effect:
- Happiness limit your growth and exansion by a lot.
That's what I'm explaining in the Ideas & Suggestions forum. Please go there instead of pursuing the discussion further. I'm explaining that GH is only a
limiting factor, not a preventing one. It's little like corruption in Civ3, you had corruption but you could fight against it. (courthouses or Communism) In the same vein you can build happiness buildings and take SPs that improve your happiness.
- Each new city increases the cost of the national wonders. There are 12 of them to build.
The only key NW is the National College. You can perfectly neglect the others.
- Grabbing too much land will upset the other civs.
It's interesting, and could be the base for a smart AI that focuses on big civs instead of attacking each others stupidly.
- Already researched techs by other civs will cost less to you.
? Is it really limiting wideness ?
- You can steal techs to more advanced players (works great at the beginning lose usefullness quickly tough).
I think I got you : while science is not so much the key, big civs are less likely to be must haves. Yes, why not, although I think those two effects are anecdotic and could be pushed more towards a true mechanic. (science diffusion would be interesting in term of flavor)
Of course there are much benefits of being in the lead, but Firaxis has been working towards a balance between tall and wide, that directly affects the snowball effect.
See the Ideas & Suggestions thread.