The whole thesis of the strongtowns.org website, with lots of data and references to back it up, is that medium dense city centers and town centers are the only parts of a city/town that are revenue positive. Commercial strip mall areas and suburbs do not generate enough sales tax and property tax revenue (and often require tax breaks) to pay for the infrastructure: miles of roads and streets, sewers, utilities, etc. Density, when you have lots of people and economic activity in a smaller area, can pay for its infrastructure, even if it might be more expensive than a country road.
Otherwise, I do agree, it is cyclical, cities were over-hyped for a while, with the tech boom in many east and west coast cities, but now remote work might reverse that. However, as Europe has shown, with a few precautions and some sacrifice such as concerts and night-clubs, a lot of city activities can go on, and with less rent pressure, we might be headed toward a good balance.
Think about it: one mile of sewer, one mile of street, supports X people in a suburb, or 50X people in a dense city.
Even if the wear costs in the city are double or even four times that, the city comes out way ahead.
https://mikethemadbiologist.com/2016/11/10/meanwhile-u-s-sub...
The last American urban exodus started in the 1950s and wasn't fully reversed until the early 2000s.
Many cities never recovered in the Mid-Atlantic and Midwest. Even the most expensive cities, like San Francisco, LA, and New York, had neighborhoods that were run down and dilapidated, ignored for decades, until very recently.
I'm not saying this will necessarily happen again, but it certainly could. Cities are incredibly expensive to maintain, and most American cities have already spent their "rainy day" funds in just a few months of lock-downs. All the while depolicing is gaining steam.