—“At what point do diseconomies of organizational scale kick in or is this a flawed frame?”—Lee Tucker

That is the entire question as far as I am concerned.
If we take an empirical look, the scale increases with available information systems, and the homogeneity of the population.

But in simple terms, a city = a market, and a market must serve a demographic and a demographic must exploit a niche in the regional or world market.

So as far as I can tell there is some ratio between population, density, iq, and homogeneity, institutiona means of distributing/collecting information necessary for decidablity, and relative productivity (purchasing power) that should tell us the optimum size.

the value of scale kicked in with gunpowder and the high cost of total war. (napoleon). HOwever, with nuclear weapons, that value disappears, so as far as I can tell, 5-10M, and that is up from what looks like 2-3M in the last century, simply because of the complexity of goods and services produced, and longer lifetimes (in other words more people does not translate to more people productiely employe) . So population can increase without affecting the systems of production.

Conversely, lets look at competition: a bigger city with more assets can endure shocks over longer periods of economic reorganization, by spending down those assets.