So the problems are non-optimal spectrum utilization and pointlessly redundant cell sites.

You could extend the idea of municipal owned last mile broadband to wireless as it is more or less the same problem. But then there would be no point in having spectrum auctions. Governments have gotten used to the money they get for awarding spectrum monopolies so I don't see the municipal wireless idea going anywhere anytime soon.

Spectrum auctions are one of the more famous results of economics theory, and the first thing that people think of when they think of Ronald Coase (who won a Nobel and is up there in the pantheon of 20th century economists).