Can Wi-Fi MANs Fly?

Over at Mobile Pipeline, they're singing a tune that I've been humming for a while. Dave Molta says that plans to offer wireless broadband citywide using WiFi are a case of "technological naiveté and ill-considered business models." Molta mentions problems ranging from the fact that WiFi was conceived as a Local solution and is thus ill-suited for wide-area use, and also that the protocol overheads reduce throughput, that interference will be immense, that there are business hurdles, and that the Municipal WiFi networks will end up like city streets - poorly maintained and full of holes. I agree with Molta overall, although I do disagree on certain points: I think that when the municipality is involved, the business model becomes much better, because they have access to infinite real estate for AP installation (ex: lampposts). Nobody has ever mentioned the possibility of using 802.11a for MANs, so let me explain why 'a' is also ill-suited: although there is currently less interference at 5GHz, it too is unlicensed and therefore interference is a problem, and can be expected to get worse; while 'a' offers more channels and faster speeds, it also has much worse range, and the number of incremental access points required would be an exponential increase from a WiFi 'b' network; and unlike 'b','a' radios are not embedded in most laptops. I'd love to be wrong. A cheap ubiquitous WiFi MAN would reduce the costs of connectivity greatly, bridge the digital divide, and make my life easier. However, there is not yet any example of a successful, loaded, WiFi MAN network anywhere in the world (networks supporting a handful of Police officers are not loaded). In fact, the case of the University of Texas shows us that interference does harm WiFi MAN.