Having said that, "web 3.0" as recently coined is the same type of middlemen-led openwashing. The real revolutionary technologies won't be hyped as some unified trend, because actual elimination of middlemen leaves no outsized investment money to pay for such hype. As always, the revolution will not be televised.
I do agree that Web 2.0 was naive in thinking that merely having open decentralized protocols was enough to ensure services built on top of them will be open and decentralized. The commercialized centralization happened despite the decentralized/open tech — on a layer above. Megacorps figured out that open protocols are merely a fertile ground for building walled gardens on.
And still around :)
edit: I'll note that if google reader were still around today, its likely the discover algorithm would have been corrupted by bots and trolls and growth seeking managers at google. But back then this hadn't happened yet so it all worked pretty good.
Similar thing happened with XMPP and Google Talk. There was a brief moment in time when you could use any instant messenger with any client of your choice (talk to both Facebook Messenger and Google Talk users via your own server). That also been killed off in favor of proprietary apps.
I'm trying to think of a current-day equivalent of what a shock and disappointment that was, but I'm afraid proprietary protocols and non-interoperable app silos became such a norm that nobody even expects things to work together. I expect GMail to drop IMAP support any day now.
We've had open Web APIs. We've had OpenID. We've had ambitions for semantic web and microformats. We've had API mash-ups. All of this has been closed, abandoned, or morphed into a locked-in monetized hell when VCs demanded their money back.
†) it's symbolic of course, but that's when shutdown of Google Reader deflated enthusiasm about syndication. About that time Google dropped their "don't be evil" motto, and wrote a spec for launching secret DRM APIs from HTML5.