Keep expanding - it's not unusual to see 100-150 cities on a 100K game.
I think that's an understatement. I think I had roughly 100 cities on
tiny. This is standard. The time and effort involved in doing one of these well is probably too great for me at the moment as I'm doing a Histographic game. But here are some tips from my last experience:
I think ToA outweighs Education by miles. It saves on maintenance and provides free temple for less pop. For the civ, I'm still undecided on who is better, Sumeria, Celts, or Babylon. AGR is very nice for the bonus food (and Marsdens idea of Warm/Dry intrigues me for requiring less workers) but the shield discounts require a lot less pop to get your buildings out earlier.
Some numbers on the big three (concering the ToA route, ignoring the initial growth of 20 food, add ten food if you wish):
Babylon = 40 sh Library and 80 sh Cath = 6 pop ~ 60 food w/ granary
Sumeria = 40 sh Library and 160 sh Cath = 10 pop ~ 100 food w/ granary
Celts = 80 sh Library and 80 sh Cath = 8 pop ~ 80 food /w granary
It's a trade off... Babylon has the least pops to sac, but more difficult in getting fast settler factories. Sumeria and Celts have the extra food for settlers and getting the extra pop they need. Celts seem to have the best of both worlds but the other two have the ability to pop out libraries at size 3 (30 food after settlement) which gets those critical building earlier while the celts will have to wait for 5 (assuming you have muskets for short rush) or 6 population to whip.
Having played Sumeria for my one game. Those 160-shield cathedrals are a pain in the rear. Even with AGR and being on Monarch, some cities were unable to get Cath early because of the happiness limitations, not to mention there is no short-rush option for 140 shields, meaning you had to get to size 4 for the last rush (120 to 160) and thus have one extra population not used for whipping. Thus, I'd probably play Celts or Babylon next time.
It's probably a close call with Celts vs. Babylon. I bet Babylon's success is a little more map dependent as it will need excellent conditions to offset the Celts growing power. The only thing Babylon has going for it in my opinion are those super early libraries (well, not entirely, free MA tech and SGL chance help some too). I also think that ToA games have been ignoring the slight maintenance edge they have over Education games to build colosseums. With their extra maintenance (free temples, not building unis) colosseums give the edge to complete a few turns sooner. (I believe this helped me get my #1 spot.) I think Babylon will take advantage of this better because they have less pop to sac and may be able to get those 6 extra pop for the colossuem earlier. Is that those enough to outweigh AGR? I don't know yet.
I agree with the earlier sentiments that 2 SGLs for pyramids and ToA will be essential for best dates. (Occasionally, an AI will build ToA for you, probably not optimal at this difficulty, but a possibility if that 2nd SGL won't pop) A third, if occurring, I would probably use for Sistine to help with happiness issues. Sometimes a core city will end up building Sistine if I don't need wealth or markets. Babylon's slightly higher SGL percentage will help reduce re-starts. Still you're giving up AGR.
Okay, those are my 2 cents on this one. I hope some people find it helpful.