This is about to change. According to CNET, Twitter's new VP of Operations, Santosh Jayaram, said that Twitter search will soon start crawling the links included in tweets. There's a lot of links there, and given Twitter's huge growth, soon these links might comprise a hefty portion of the overall web, making it a much more complete search engine that it currently is. It's a big technical leap, but it shows that the folks at Twitter are serious about search.
Jayaram gave another very interesting hint at where they're going: Twitter Search will also get a reputation ranking system. That means that not all tweets will be equal; rank will be calculated for each twitterer, probably based on several criteria such as number of followers, number of retweets and so forth.
When I recently wrote about Tweefind, a Twitter search that also takes rank into account, some commenters asked why are we putting so much emphasis on a very beta, buggy service. Well, regardless of the actual implementation, it was immediately obvious that the idea is not only great; it's precisely what Twitter needs.
This other part of Twitter's search strategy is equally important, and together these might prove to be the one-two punch they need to put some fear into their competitors. Crawling links will give Twitter's search engine the scope it needs to compete with Google. Rank will give it a way to make sense out of all that data and decide what's important.