It seems to me that Frank Rizzo’s reasoning in deciding that government-run broadband Internet is a bad idea applies pretty much across the board for possible government actions beyond a limited set of activities:

At the heart of the problem is this: The economics simply didn’t work [in Philadelphia]. To come close to breaking even, municipal systems need to attract sufficient numbers of low-dollar subscribers to help offset the ever-swelling capital costs of building, maintaining and upgrading the network.
Typically, any wire line or wireless broadband network will cost, conservatively, tens of millions of dollars in initial investments. On top of massive start-up capital costs for initial construction, broadband networks require huge annual operating costs to pay for administrative staff, customer service, repairs and maintenance. Equipment upgrades — needed every four to five years — often cost potentially tens of millions of dollars more.
To offset these costs, municipal systems need to attract thousands of local subscribers by either drawing customers away from commercial providers or by persuading nonbroadband users to sign up.
But commercial providers generally offer more reliable and faster service — few of their subscribers are likely to switch to a slower municipal service to save a couple of bucks. And, as the Pew Internet & American Life Project has found, broadband nonusers don’t see relevance of the technology in their lives, making it unlikely that a taxpayer-subsidized network would suddenly change their minds.

Government isn’t as sufficient. It can rig the system. And it drives up prices for everybody outside of its offering and diminishes quality for those within.
Despite his insight, Rizzo falls back on brainstorming ways in which to make the system work:

What’s really needed is not a utopian dream bound for fiscal bankruptcy, but rather a true national broadband policy that will give the nation’s mayors the resources for low-cost computers, digital training, local technology centers and resources for creative nonprofits and other third parties to generate targeted online content that will foster greater interest in broadband by nonusers. When Wireless Philadelphia failed, we did just this with the Digital Inclusion 2.0 program and, as a result, more low-income residents are online in our city than ever before.

Thus does a limited effort to level the playing field and create a baseline public infrastructure for Internet access grows into the generation of content meant to generate interest in a government service. It’s like government mission-creep at Internet speed.