The problem with it is economics of scale. If a city has to buy trucks, secondary market fiber optics, train employees, buy routers and switches every five years, purchase boring equipment, and additional management to get it to work with other utilities the city offers, it is going to be extremely expensive per user per month. A lot of cities who want to do this want to tie this in with the electical departments, but electrical is a huge moneymaker, ISPs never will be on a small scale. It is a service at a loss. It can’t replace cable TV right now (yet).
The good news is that there is a few wireless protocols on the horizon that look very promising for delivering large amounts of data long distances and with better home penetration. But that stuff is always 5–10 years away from commercial viablity, and never is 100% effective.