Dabble has been around for a while. They have a new video, Digg into Dabble, showing how they scrape Digg.com and get at all the underlying data to do some stats analysis. Unix heads will love the pipes and filters Dabble replace a lot of time spent at the command line. I appreciated this quote from the comments:

Today when we query a database we get a recordset.
Tomorrow when we query a database we will get another database.

Dabble enables anyone to scrape various feeds and data in tables, parse, analyze, manipulate and re-publish for others to consume. It would be cool to turn Myspace into an RSS feed or a giant database with a well-documented and powerful API. Think of the cool widgets people could make.