Well, the point was that decentralized internets would probably feel like the early internet did at first, but if I think about it now, that wouldn't really be the case. It'd probably be a lot like the current internet, and you probably wouldn't even notice the change.
It builds up on many of the same technologies that the current internet is built upon, so it would bring very little new stuff to the table.
What would really be neat is something using technologies that people haven't figured out yet. The web is a boring and "safe" place by now because everything seems like it's already been tried out, and nothing feels new. All of this contributes to making the web more appealing to the general public, which is a double-edged sword; on one hand, more people might bring new perspectives on all matters; on the other, groups are at their best when they're small, as further explained by >>155
I really don't know if you could make something out of VR/AR, which right now only feels like an expensive gimmick made to sell expensive hardware, or maybe something else entirely, but it should probably be something that makes people interact with each other in ways that they've never experienced before, and that encourages emergent behaviors.
On this topic, I hope you'll forgive me if it's not exactly text, but I'd like to recommend a lecture by Brian Moriarty, one of the authors of some of the best adventure games of the 80s and early 90s. It's mostly about videogames, but also partly about how environments like MMOs and the internet itself (used to) encourage emergent behavior, and it's also just a very good lecture by itself:http://www.gdcvault.com/play/1016390/Listen-The-Potential-of-Shared
(On a side note, you can easily play this with streamlink if you'd rather not use flash.)
Here's the transcript, if you'd rather just read, but it's definitely better to listen!http://ludix.com/moriarty/listen.html