Broadly speaking, though, I think you’re missing the point on maybe two levels.

First level. I resist comparison to libraries. News mostly isn’t libraries. News mostly isn’t scientific research. Which is not to say that news–or what I’ll call “news,” for lack of a better term–is vapid gossip. It ranges between those two extremes. And in that range, I believe we care about what people say and what is said abut them. It’s conversational. See here for background on how a parable of the fishing village can help us reorient ourselves to what’s fundamentally important about the news:

http://networkednews.wordpress.com/2009/04/03/thinking-the-unthinkable-parable-of-the-future-of-news/

My sense is that news is a social thing, a stratum of information in which people form groups, identities, and other bonds. Now, is it all conversation? No, of course not. It’s often research, like that done in a library. But news–again, very broadly speaking–has a special kind of social life that obtains between people around the water cooler or on twitter. That social life is horizontal; it’s one-to-one and one-to-many, very many times over.

Second level. I agree that you don’t want, and can’t handle all of (3)–which, if you recall, comprises health care news that others have shared with Robert Reich, whether via twitter or otherwise. Fine, but you’re (understandably) taking me to be saying more than I really am (don’t worry, I ain’t mad).

So let me be clear(er). I think you do want the health care news to which my tweets that mention Robert Reich point. I think you do want the health care news to which Paul Krugman’s tweets that mention Robert Reich point. In other words, you care about some of (3).

If there were a good way to establish expertise or “trustedness” on a topic (like health care or medicine or politics), you might want health care news to which such experts’ tweets that mention Robert Reich point. And I do think there’s a good way to establish such expertise. In fact, this very system of following does just that. (See tunkrank for background: http://thenoisychannel.com/2009/01/13/a-twitter-analog-to-pagerank/. You don’t lack the math chops.)

The fact that someone reveals a preference for health care news from Paul Krugman indicates that someone trusts Krugman on health care. If you repeat that many times over–i.e., with enough people independently following users A, B, and C on topics X, Y, and Z–you’ve got socially networked content recommendation engine. And that’s pretty cool–far-out, sure, but cool, imho.

I hope this helps a bit!

The idea that (1), (2) and (3) are all the same seems to me a lot like saying that library search engines should not ask you specify whether you want to search “by author” or ‘by subject.” But that specification is super (even perhaps plus duper) helpful? If I want articles about the awards Paul Krugman has won, it’s less helpful to run a search that also pulls up all the articles Krugman himself has written. Likewise, if I want to research what Krugman’s written, it’s not helpful to run a search that also pulls up every article that mentions him in any way.