OK, here goes...

First, I agree that there's far more to be technically wrung out of the PSTN -- although I'm dubious we'll see much of it. iobi was probably the last gasp, bar the FMC bust of activity that most users won't care about very much. Delivering a bit of goodness to lots of people is every bit as honorable as delivering a ton to a small set of folk.

The problems of the PSTN run deep. By bolting connectivity to service, operators tend to be geographically aligned. My wife and her parents don't get the same service provider because they don't life in the same country. A market-aligned provider will be able to target new functionality to people who have some need in common. The PSTN is more like a set of supermarkets -- the first catering to people who came in Toyotas or Lexuses, the second for Fords, Volvos and Jaguars, etc. The users don't have enough in common to make any one new feature have enough mass appeal.

Also, the moment any new feature requires any new sort of signalling or media, or involves a 3rd party, it's game over for innovation from the edge. The examples listed are pretty much all you'll ever see. We've mined out the technical envelope of pure 'edge innovation' for the PSTN.

That's probably why so many of us seceeded at the first chance to Skype, which gave us more of the functions we wanted in one simple, integrated, free package.