"Every city beyond the first causes you to shift work to commerce tiles in order to keep up research. AND cottages take forever to pay off."
That kinda makes my head hurt.
/pretty soon we will all be starting over. I might actually read the manual this time.
I didn't figure this out until I spent a long time trying to answer, how many food/hammers is 1 commerce worth early game?
To use a simple example, say your 3rd city increases your total maintenance costs (which is more than it's own maintenance cost) 4 gold per turn. If you want to maintain your prior beakers per turn, you will have to come up with 4 commerce. In most cases, this means taking two population off of mines (4 food+hammers) in favor of riverside cottage (2 food + 2 commerce). 4 food + hammers - 2 food = 2 food+hammers.
Therefore, settling that new city actually cost you 2 x 2 food+hammers = 4 food + hammers per turn, making that unirrigated rice you gained a lot less impressive.
Why not just keep the mines and ignore the gold loss? If you've ever stalled your economy before you hit pottery/writing, you'll realize it can cost you the game techwise. A seemingly meaningless decision now depends on the AIs tech rate and how close you are to matching it.
About the expand to 0% then cottage: through experience shadowing games at various levels, I've come to realize that this advice implicitly relies on lower level maintenance discounts/tech discounts. Specifically, if you hit 0% before you hit pottery/writing, you lose, but on say noble/prince/monarch tech discounts will allow you to reach those before you can expand fast enough to kill your economy.
Second, each difficulty level below deity gains about a 5% maintenance discount, so on noble, you get about a 25% maintenance discount. Which means you'll have to switch off 25% fewer tiles to commerce, which means new cities are 25% x maintenance cost more productive than in deity. And more commerce x lower research cost means you hit those economic techs a lot faster.
Finally let me point out that due to optimization in difficulty adjustments, the best strategies for deity are somewhat suboptimal for say monarch, and I'm not just talking warrior rushes. Higher levels require a lot of suboptimal sacrifices, but in say monarch you can adhere to land is power without falling behind the AIs tech rate.