I just had the silliest thought. Think of it as a step beyond vanity searching on Google.
Say you ran your own Googlebot-esque browser, and you somehow thresholded it so that you only got the top results -- so that you had megabytes or gigabytes of search results rather than terabytes. IIRC, Google uses a small number of "root sites" which get all the initial "Google juice" -- the initial search importance. It then calculates the spillover of that importance to other sites. So you could do the same thing, but use different sites as the roots. If you did it with, say, your own personal site, your LiveJournal, and maybe some friends sites as the roots, you could get a "me Google", sorta. So I could do a "Noah Search" to see whether a particular concept had any relevance to me, as opposed to regular Google, which tells whether it has relevance to the web at large. A search on "monkey speak" would definitely come up with Tom's site as the first hit, for instance, since he'd have such a lot of "NoahGoogle juice" :-)
That'd be silly. It would also take a very large amount of computational resources. I like the concept, though.