No, it is not anecdotal. I’ve linked to this article by Ric Stern. He’s the inventor of the ramp test and according to his blog post, 97 % of the athletes whose data he has had access to had their FTP fall within 72 and 77 % of MAP. Perhaps there is newer research out there, but that’s where I got my numbers from. He wrote that he was contacted by TR and they asked him, and back then he (in his own words) lazily replied 75 % (which is roughly in the middle of 72 % and 77 %). He now recommends a slightly different approach for inexperienced people using a ramp test alone, namely that they start at the low end.