There might really be the magic window, where a place is big enough for real opportunities and amenities but small enough for cheap large-lot housing, high-mileage low-traffic commutes, and ample free parking. But if it exists, it's very narrow. We have no track record of getting population centers into that window or holding them there. Most population centers, whether driven by government spending or not, either undershoot or overshoot it in the end.

We have thousands upon thousands of towns sustained by the local military installation, prison, defense plant, research lab, state university, etc. Even some mid-sized metro areas where the economy is mostly Medicare. Usually it is more of a life support thing. They're too small to be desirable. On those occasions when they do take off and become desirable (defense spending made Silicon Valley!) they quickly become competitive.

Rather than trying to manufacture and maintain dozens of places on this knife's edge Goldilocks window for postwar sprawl (that may or may not exist), we could instead choose a building and transportation architecture that provides a high quality of life at reasonable cost over a wider range of population levels.