I think one thing to be said here is that Only a Couple CITIES (with the really good buildings) plus a spamming of tiny settlements (outposts of 50-100 population or less) is kind of what we WANT. yes?
The more CITIES you have, the far more research and production and prestige your empire will attain. The BIGGER your cities are (more pop) the more research, production, and prestige. Its not a linear factor, 2 cities of 250 citizens are better than 5 cities of 100 citizens. 1 City of 500 population should be EVEN BETTER than 2 cities of 250 pop.
The food would limit your City population, and I propose all new settlements are Guaranteed 10-50 population based upon terrain (maybe a diff number, but according to the example it makes sense). In this way, you can spam as many CRAP outposts as you want with 50 citizens. They will do hardly anythihng for you other than expand territory and resource aquisition, they will not be able to reasearch or produce things worth CRAP. they will take 15 turns to build one soldier with ARMOR!!!! while your big city can build 1000 legionairres in maybe 5 turns. The amount your cities can grow beyond pop 50 will be determined by available food, and those cities with higher proportions of Prestige will get a larger proportion of that population.
youll want your outposts to have 0 prestige in order to not detract from the population of your Bread-Winner cities. In fact, you could create settlements to Harvest more food, simply for Food = Population sake, and solely to ship all that food for your big cities. Having 3 outposts of 50 citizens each, farm a strip of fertile land worth 1000 food, then your 1 city of 500 can now grow to 1500 (or 1000, or any 500 + divisible of 1000). Would it be most efficient to move all your food into one city?? Yes. Up until the point where your Big-City's housing can no longer hold population, in this case, excess food would be distributed to other settlements, first based on prestige, second based on distance to BigCity, and third based on distance from food's origin OR that being too complicated, third could simply be the natural fertility of the area (which is SORT OF like distance from food's origin).
In this way, say City A can hold 1000 citizens, and you have 500, and you have food enough for another 1000 citizens. Your city maxes growth at 1000, (500 additional people), and then City B which is pop 100 and can house 500, which has second highest prestige, gets 400 of growth, and then the last 100 are distributed among the smaller settlements according to equation (the one I colloqually spelled out, counting prestige first, distance second, ect ... lets assume that all other cities are only resource gathering settlements, and therefore its based upon distance from City
.
Then your City A upgrades to City size 3, and can now build more buildings. City A can either build more housing (+500 citizen cap) research better Houses (+1000 citizen cap) or both (+2000 citizen cap), as the technology would double the house's effieciency. So lets say they tech for better houses only, and do other interesting things with the new buildable land. Its cap is now 2000, and it grows to 1500, reducing City B back to 100 pop. Perhaps you can control this with Governors, perhaps all citizens leak into highest prestige first (like the example) or perhaps the Populations of City A and City B reach an Equilibrium based upon relative Prestige levels.
I personally like the Equilibrium idea, although the Totalitarian Prestige idea presented would also work, since the biggest City would always be the most efficient location for population, and all settlements are guaranteed at least 50 pop. On the Equilibrium idea, lets say City A had 1000 prestige, and City B had 500 prestige. I think the equilibrium could work on either a direct proportional corrolation or an exponential porportional correlation or a logrithmic porportional correlation.
I personally prefer a Direct Proportional corrolation because its easier to plan and work with, and less guess work (although you could get used to either method). In this case, the City A's growth "soft cap" would center at twice the pop of City B. So, in the most simplistic assumption (assuming there is PLENTY of living space, and a deficiency of food), lets reapproach the initial setup. City A would not have 500 (+50) pop, and City B would not have 100 (+50) pop. Instead City A would have 400 (+50) and City B would have 200 (+50). If enough food for 1000 new people were added, then City A would rise to 1067 (+50) pop and City B would rise to 533 (+50) pop. This is my preferred system, although I suppose that even if such a system were not held to, it at least reveals additional ways to PREFER isolated Metropolises supported by MANY small towns and to NEGATE the view that many smaller towns are better. In addition your smaller town's cannot individually grow into larger towns, each population center would need to be supported by farming, and in the most basic instance of infinite accesibility and resource, the player would pour all available population into one Super-Mega city, as a result of this city being the only city in the empire with Prestige.
This method uses part of the All-father of prestige as a tool for Populatoin Growth-rates and Population Distribution, however only withihn the context of Population Capacity. Population Capacity comes first, through housing and farms, and population growth fills them into their perspective roles. Prestige, in this function, would not be a factor in determining Capacity for Population, only the speed at which it attains its capacity. In this way, I think any growth at all should require at least some prestige. As you can see in this example, City A grows twice as fast as City B, however City A can only have double the amount of population as City B. Both of these factors stem directly from the amount of Prestige, and incidentally all newly growing cities will achieve max population at the same time.
**as a note of interest, this Metropolis-focused plan of national growth reminds me *slightly* of the Kuriotates from FFH, although the Kuriotates were forced into such a device by arbitrary restrictions, rather than actual reasons. In my creation of Viewpoint, larger cities are so much better that its far Superior to have less, bigger cities. This method DOES NOT require much in the way of distance maintanence in the smaller outposts to prevent city spam, as its not city spam but village/outpost spam in the fight for resources, which are very poorly defensible, and quite juicy prey to raiders. They would most likely be the first target in warfare, and there would most likely be a literal *sea* of such small settlements, where-ever there is develop-able land with some pertinent resource.