I'm all for this, and consider it to be inevitable in the long run. In the short term, however, it seems like the major hurdle will be getting one of these projects into the mainstream: for the most part, the web already does what most people want it to do, and those people aren't going to be bothered to install a new web browser so that they can do things they're already doing. Especially if it lacks the features, performance, or ease-of-use of their current browser.

So, how do we address this? Is there a "killer app" for the distributed web that will motivate people to move to it? Can we use existing web tech like Web-RTC to bootstrap the system? Maybe a workable avenue is mobile, where people are pretty comfortable installing new applications - what if we built the next social network into an app based on the distributed web?

I don't know the answer, but I'd love to hear any ideas/brainstorming you clever people have to offer.