I'm not familiar with Tweetscan's specific implementation, but I would assume they're just grabbing the public XML feed, no?