Twitscoop was just profiled over at Techcrunch by Leena Rao, and quite frankly even though the site is so new it’s Alpha edges are showing, it’s the closest thing to the “real-time web” memetracker I’ve always wanted in my pocket.

The graphs on current trending topics are pretty eye-candy for breaking up the page maintaining a good set of eyelines, but the killer app is the live-updating tag cloud to the right.

I’m not sure that I’ve ever discussed my idea for the “real-time tag cloud” publicly before, but I’ve spoken about it privately to a number of folks.

A tag cloud that dynamically and constantly updates based off not just Twitter, but other real-time content sources like Yahoo Buzz, Digg, Flickr, FriendFeed (and whenever they finally allow it, Facebook) that could sit on a touch screen like device like the Chumby.

Rather than having pop-up or click actions that simply provide the source material, in my imaginary device when you clicked or touched a keyword, the aggregated links and media resources being mentioned in the scanned stream would be emailed to me or placed in an RSS feed I subscribe to.

The goal is to stay on top of emerging trends and detecting them before they become overwhelming tsunamis.

This is useful to me as someone who makes his living off the news-cycle, but it’s also beneficial for large brands that are looking to prevent a “mass fail event” (see: #AmazonFail).

This general area of the web is fertile ground for developers with an aptitude for algorithms, since most of what exists in the space is a matter of simple keyword matching. Very few players are doing intelligent things with that keyword-matched data, but it looks like Twitscoop is working on it in an intelligent way.