1) Why is a second city being too far away from the capital a bad thing? Because of maintenance costs? Or because it takes too much time and too many worker turns to build roads connecting the cities?
Both. But maintenance costs are not that bad on the lower levels, so you can get away with settling further away as well. But if you for example are going for a HA rush, then spending 10 turns with your worker building roads will slow down your attack date a lot.
2) I actually ruled out 2N of marble when I tried to find the candidate places for the second city because its workable area would have overlapping tiles with the capital (6 tiles would overlap if I don't miscount). Isn't that a bad thing?
Overlap is good, especially with your capital. The reason is cottages. I'll try to explain...
Why cottages? Your capital gets a +50% commerce boost from Bureaucracy, and usually you aim to build an academy in the capital with a Great Scientist for +50% research. Academy is best in capital, because of the commerce boost from Bureau. Add in a library and an university and you have +100% boost to research (50%+25%+25%). By the time you get Printing Press each town will produce 6
, this is multiplied by 1.5 from bureau and becomes 9
, which goes through library, university and academy for +100% boost and becomes 18
/turn. That's from a single town. I mentioned that a decent mid-late game cap has 10 towns. That's 180
/turn, add in the commerce from palace, trade routes and other tiles you are working and it easily becomes 250
/turn with slider at 100%. Throw in Oxford university and now you are making 375
/turn in your capital alone.
Now look back to the 920AD screenshot you posted in the beginning of this thread. With slider at 50%, your entire empire produces <100
/turn, so <200 beakers if you turn up slider to max. Your capital alone could make twice as much with properly developed cottages and Oxford. Now you might be pointing out that you didn't have Printing Press and Education by that date, but in case your capital is improved properly, you are researching so much faster that you would have. Players like
Seraiel can routinely have Oxford built in the BCs. That takes quite a lot of practice (and a strong starting location), but having it before 1000 AD is very doable even for the more novice player. (Note that Oxford is by no means required in every game, often it is better to just put those hammers into military units and kill everybody.)
So, how do you get 10 or more towns in capital? A cottage must be worked by the city to grow. On normal speed Cottage->Hamlet takes 10 turns, Hamlet->Village takes 20 turns and Village->Town takes 40 turns, total time for cottage to become a town is 70 turns. Leaving it up to the capital to grow 10+ cottages would mean that it can start working the last of them very late. It takes long to get happy cap up, you need to whip some basic infra and maybe a settler early and the city must grow.
This is where helper cities come in. 2N of marble can work 4 tiles in the capital's BFC that should be cottaged (FP + 3 riverside grassland tiles). The capital has 6 other tiles that are good for cottaging (grasslands and floodplains). So until the capital grows past size 8 (food+6 cottages), the city north of marble can work and develop 4 cottages that the capital can work later. After that, every time the capital grows it will get a cottage that is already developed into at least a hamlet, maybe even a village already, which means that the new pop in capital immediately gets a great boost from bureau and academy. Without helper cities, you would always start with a new cottage from 0. By the time the capital is big enough to work all cottages, they can all already be villages and you are 30+ turns closer to towns.
Sometimes it might make sense to settle even more helper cities, if there is a lot of cottageable land. I should point out that the setup here is not ideal, because the helper city would get riverside cottages while the capital has a few non riverside cottages that produce one less commerce. Because of bureau and academy bonus in capital, it's better if capital can work the strongest cottages and helper cities can grow the non riverside cottages before they are assigned to the capital. If possible.
Hope that makes at least some sense...
There are other reasons why overlap is good also. Sharing resources is nice, for example. Especially if you have a lot of food in some city. Sometimes you have so much that you cannot work them all without growing into unhappiness, then you can borrow some to another city, and borrow them back if you have to regrow quickly after a whip. Close cities also makes worker logistics a lot easier.
. That's not optimal.
, you'd need to work 3 unimproved Floodplains to get the same!
