Some strategies for discovering things that go unnoticed would involve ignoring the “junk mail” and go to the (usually smaller) shops in the side streets. These often offer better deals because they do not sell low margin junk and compensate for it by volume and advertising like the mass stores. Yesterday I alluded to the fact that search engines may have begun to suffer from the same problem by ranking according to popularity or by being gamed by specific writing styles (aka “marketing” in the real world). In other words, search engines show the main streets, first and foremost, and not so much the side streets. To wit, a search engine would show a main street shop if it had just one item of what you were asking for instead of showing a side street shop with more selections. New shops in particular have a huge problem since they are judged by the network rather than their inherent quality.
One way of discovering new blogs is to use other blogs as a reference. This, I think, does not work nearly as well as it could.
I do not think a blog roll is any substitute for that. Blogroll often turn into a mutual link exchanges (I thought of a different word for it, but it wasn’t fit for publication ), consequently grow huge, etc. They are often not reflective of what the bloggers actually read.
Maybe it is time to return to the roots: A blog is really short for weblog and in a blog one is supposed to write about what one did on the web on that particular day. It may be as simple as a rough edit of one’s browser history.