>However, as cool as this is, it's just another way to "scrape"

Isn't that the point? The demo seems like it'd be a lot easier, less verbose, and probably less brittle, than using cUrl/xpaths or otherwise parsing that HTML yourself.