The big difference at this point in time, as best I can tell, is Twine's ability to auto-discover relevant content tags for you when you submit links -- tags I

The big difference at this point in time, as best I can tell, is Twine's ability to auto-discover relevant content tags for you when you submit links -- tags I might not have entered or certainly that I wouldn't have taken the time to enter since I never enter more than four or five tags but it is not uncommon for twine to discover a dozen or more. Given that richer set of tags, it is more likely that it can do a better job finding related content and interesting linkages between content. I think Twine, as it exists now, is really an app to feed their internal engine with interesting data that they'll one day use to reveal their real application and it's not to difficult to imagine that such an application might involve a switch from passive data collection to active data discovery with intelligent summarization, but that's just my best guess.