Scoopler is indexing live streams from Twitter, Digg, Delicious, Flickr and Identica in real-time, and not only that; it's also indexing links, videos and photos from these data sources, which is similar to what Twitter has been planning to do. As a result, Scoopler doesn't feel so much as a search engine, more like a news site with a constant stream of live updates in the middle (paused when you mouseover, a nice touch), list of hot topics on the left, and most popular results on the right.
These popular results, based on the number of shares for certain item (remember, Scoopler is indexing social sites, where each item has a social context; on Twitter it can be retweets, on Digg it can be diggs, etc.) are exactly what Twitter search is missing. Unfortunately, Scoopler doesn't take the concept too far; you can only filter hot results by content (links, images, videos), but you cannot sort them in any way. Furthermore, the way Scoopler calculates which results are hot is questionable; as I've written earlier, calculating rank or reputation for real time items can be a very complex task, and Scoopler seems to take only the number of shares for each item into account.
Still, Scoopler shows promise. It's fast, quite elegant (the Peek feature lets you preview an item in an overlay window, which works surprisingly well, even for videos), and, most importantly, it showed up at the right time, when real-time search is a very hot topic. If the team behind it keeps tweaking and improving the engine (they promise to add more data sources soon; they need them, because right now the results are Twitter-infested), it might turn into a search force to be reckoned with.