What do you suppose would happen if we all visited these URLs from every public access point we use? (edit: and shared VPNs, Tor, corporate networks, etc.)
If I was building an unruly scraper (which, in case our new overlords are listening, I would never do), I would read robots.txt so that I had a clue where the secret information that the company did not want me to read was located.
I'm not allowed to look in /documents/source/? Perfect. Let's start there.
They might be using it as a way to find specific pages which have content Yelp doesn't want you to scrape. The "evil" scenario.
They might also not be looking at the file, and just appending random words to the end of Yelp's biz URLs to scrape every business. Which, at some point, it might hit that URL by accident, since they are all words you'd find in other business URLs. Though this seems less likely.
Scrapers can be pretty shady, so they're isn't a good way to ensure that web traffic is coming from a legitimate human using a browser. To the server it's all just bits on the wire.