It's simply checking if the connection is from an actual browser, as a scraper pretending to be one won't actually refresh the page as instructed. It's going to buy some time, but like the rest of Anubis in general, it will only work until the scrapers get modified to work around it.