After reading the "Saving the Net" link via Julian, and after spending the last few weeks in dial-up hell with no end in sight, I can report, without fear of contradiction, this about broadband in America:
1) It sucks.
2) Further, it is likely to go on sucking until such a time as the FCC and all the various state and local regulatory bodies responsible for the suckage are ground into dust. No halfway measure or incremental change is possible. We've been there and tried that, and the regulatory system will always game itself against innovators and consumers.
3) The current politico-regulatory-business model, i.e. fascist plan, for broadband provides nearly zero incentive to support and maintain existing and soon-to-be built infrastructure. The assumption is that once reached and enticed with a long list of "new" features, consumers will be held relatively captive with little ability to move freely between competing, equivalent services. Hence, there is no reason to invest in actually providing, 24/7, the service consumers contract for—they have no where else better to go and the marginal cost of switching is a powerful disincentive to even try.
4) A chunk of the resulting duopoly profits is expected to be routed back to the political class and their favored projects, a shakedown of immense proportions. If subsidies to certain classes of end-users of telecom are deemed absolutely necessary by the State– and they will be—then only the end-users should receive them. Such "phone stamps" would deliver the proscribed service without totally warping the telecom market and propping up decrepit providers, and would advance the principle that consumers should choose their info providers, and not the other way around.
5) Such change, while absolutely essentially to avoid locking much of the U.S. into a needlessly costly and complicated state-of-the-art circa 1933 info-regime, is about as likely as Time Warner finding and fixing the problem on node 61, trunk C41, head 20 before the sun rises on the morrow.