The New York Times runs this story somewhat later than one might expect: I wrote a similar story that appeared last August in the Economist. This Times piece focused a bit more on digital divide issues without, of course, providing any substantiation that putting computers in people's homes makes a difference in their lives. Sure, they have a great anecdote here that a student with a computer and Wi-Fi service doesn't have to take a bus in a dangerous neighborhood at night to get to a library; but that hardly equates to information that kids and adults see benefits in their lives, like higher income, less drug use, higher entry into education, and less unemployment.
The article notes, "Philadelphia officials say service will not be disconnected." This is highly uncertain. They may get some cancellation penalties from EarthLink, and the $4m estimated to complete the network is both specious (the last part of a network is always more expensive to build; it's not linear to get to completion), and doesn't talk about the millions in annual operating cost. No private operator would take this except under contract.
Philadelphia's current CIO is noted as saying that "[m]arketing was also slow to begin, so paid subscribers did not sign up in the numbers that providers initially hoped." Also specious. Without a network that worked well, EarthLink wasn't inclined to market heavily. The same is true in most early big-city networks. Service wasn't good; why advertise for users?
The article makes the good point that the cost of broadband has dropped (at least at entry-level points) over the last three years, making cheap Wi-Fi less of a draw than it was in late 2004.
One major error in this piece, not uncommon: "Unlike most other cities where municipal wireless was going to be offered in free hotspots and at a reduced price for residential service, San Francisco planned to offer citywide wireless free in a three-way deal with EarthLink, which was to build the grid, and Google, which would have paid to advertise through the network." No. Google was going to subsidize a slower service, and EarthLink was offering a paid faster service. Google's service might not have featured any advertising, either. Further, MetroFi's model, in operation in several cities, offers free, ad-supported service, too. They're only mentioned in passing and not by name.
Sascha Meinrath is also quoted in this article using old numbers (Sept. 2006) about St. Cloud, Florida's network, a free service built at city expense. Those numbers are now 18 months old, weren't provided by the city, and the mayor who inherited the already-built system disputed some of the statistics and was vocal about various problems with the network.