Surely this can't be too hard to design and implement - most likely the impossible challenge is the chicken-and-egg problem - how can you get people to use it.
I think what's missing is a compelling user experience that's decentralized by its nature. Even the terminology needs to be thought out:
email : {Gmail, Hotmail, ...} :: ____ : {Twitter, Facebook, ...}
maybe ____ is "stream" as in "what's your stream address?"
The technology side is exciting, but the UX needs to be nailed before bothering with it. As for adoption, there are a ton of social networking clients out there, I'm sure you could convince a few to adopt this, maybe by starting with a niche first.
But you're right - we need to nail the user experience.
I would argue that it doesn't.
Twitter, in the social sense, has value for users on both sides as a corral of content consumers and producers. Twitter, in the technical sense, offers very little - the only special consideration is that I can send an SMS message from my phone and it will appear on a website, but there are many other services that provide this functionality. The reason I use Twitter is that it's free.
Short answer: yes. The communication model/platform has viability, but implementation(s) would be what makes or breaks it.
Obvious I know, but it bears repeating.