Wow, if managing 20 tabs is a problem for these guys, they'll panic in front of my 200+ tabs sessions
I'm interested in what will come out of this, even if some of the articles they link practically solve many of the inherent problems of tabs without reinventing the wheel or doing weird things (I particularly like Raskin's mockup and the solutions he proposes). I hope the ideas people submit don't try to solve everything they can in a single interface (like Mahemoff apparently suggests), otherwise they'll end up with complex interfaces that try to anticipate every user needs and will only confuse the hell out of the user.
Personally, I'd like to see something like Internet Explorer QuickTabs or OS X Exposé on steroids. Instead of having a tab bar, you would have a thumbnail page with realtime previews of the pages you have open (taking advantage of GPU acceleration, which at least Opera Software is working on). You can move them around, group them if you like, have various thumb pages to overcome screen space limitations, the browser can notify you of specific pages incidents via visual cues on its thumbnail, etc. Clicking on one would of course take you to that page, and you could easily go back to the thumbs with a hotkey or hovering the mouse over a previously defined hotcorner.
That said, I personally think that, for most people, all of these are solutions in search of a problem, and they don't present clear advantages to them. Not to mention that, one way or another, browsers already have methods to solve many of the problems outlined. The real problems, not the "things I'd like to have", as those could be cool for a while and remain unused for the rest of the time.