I'm definitely not the first person to comment on this, in fact, Topsy has been working on this principle for 3 years, but Twitter's changes are a step towards ensuring that one of the most powerful features of Twitter (search) stays in house.

One of the main problems with Twitter Search is that it has little to no "rank" or reputation information - that is, a nameless-nobody spamming the twitter stream will come up just as readily in search as a high profile celebrity or commentator. Whilst this may be desirable in some cases, it seriously limits the usefulness of the search feature in terms of seeing what is going on that actually matters.

Google's famous algorithm, PageRank, solves this partly by using inbound links to build credibility, ie: the more people who link to you, the more important you are. Twitter has always had this to some degree, in that the number of followers could in theory be used to denote how important a given individual may be, but as we've all seen recently, it's way too easy to get lots of useless/fake followers through the use of various spam/follow-bots/etc.

Enter the retweet API - All of a sudden, the amount of social worth can be tracked by how often your tweet gets re-posted, the theory being that the more valuable what you have to say is, the more often it will be retweeted, hence building your reputation or "social worth". As the TechCrunch article (and many others) have said: "Retweeting becomes the currency of the web".

With social rank (TweetRank?) being considered, all of a sudden Twitter (or other realtime search engines) can return results based on importance rather than just recency. Being at the top of realtime search results is quickly becoming potentially as valuable as coming on the first page of Google (imagine searching for "best restaurant" and getting nearby recommendations based on rank rather than recency).

Similarly, the retweet API opens up a whole new range of twitter abuse (ie: auto-retweet bots, retweet baiting and much, much more), so it's going to be interesting to see how this plays out. I'll be interested to see how much of the power of the API is opened up to third parties with respect to search - ie: will Topsy/Google get access to this data or is that portion going to remain in-house?