There is nothing of real documentation pursuant to how to use the crawl feature. It would be awesome to know how to utilize it. I would suggest a test run with a bunk account. You may easily discover it's use, but from the limited documentation I can confirm it CAN be used to scrape tweets or other "information" from other pages and feeds. This can include a twitter account, but I think it was generally more intended to use with feeds where you might grab certain content from within a feed such as an image, or a product description, etc... It's basically a super weak and limited scraping tool, but I'm sure someone knows how to use it properly.
I suspect it's a very powerful overall feature, and one that with proper utilization can make all the difference in the effectiveness of twitter campaigns. I find FL to be generally very lacking in customization. While it does have a number of great features it can be quite challenging and convoluted to put an actual project together. I hope the software continue to improve.