Wide Area WiFi Is a Pipe Dream
The dream of ubiquitous 802.11 wireless access for all is one that lives in the minds of idealists and marketing types. Idealists can be excused, because hot-spots will always be a useful resource (assuming you can wean the UK police off their habit of arresting anyone using a computer in public on the charge of 'stealing broadband access'). Marketing types cannot be excused -- all they saw was huge orders for equipment and endless streams of revenue, and their greed made them deaf to people who might raise issues about the technology.
The problem's the protocol. Wireless is not about data and data rates, its all about chopping up time into pieces. The 802.11 protocol chops time into quite generous chunks, especially if several stations are contending for the link, and this limits the amount of users that can be accommodated by each link. In other words, it works fine at home (most of the time) and in the office where traffic is spasmodic, but put a whole bunch of users on the link and/or try streaming some data and performance goes down the tubes big time. Various strategems have been proposed to overcome the problems with the protocol, many of them very ingenious, but ultimately you're always running up against the problem of trying to use a protocol for a purpose that it wasn't designed for. Given enough effort and equipment you may well get a system to work, but you may find that what you end up with is nothing like the network of low cost access points that was in the original Powerpoint slides.
These problems may explain why over the last year or so the vehement and concerted opposition to municipal WiFi networks by established wireless providers seems to have just faded away. They obviously knew something was up.