My main issue with the Mayfield survey was not the rating itself - I don't necessarily think there was EF scale 'abuse' in the sense blatant misrating occurred. Maybe some of the ratings were slightly on the low end. For that "awesomely built home" I could see an argument for 200mph (Like Rochelle) but that was maybe avoided for being too close to an EF5 and the fact past EF5s have been rated 200mph. The 200/EF4 rating seems almost 'retired' nowadays. But if single homes can't receive EF5 damage unless exceptionally well built (which seems to be the current way of things despite that not always being the situation), then I don't think PAH messed up in that sense.
Rather, my main issue was the rushed/incomplete nature of the survey. I'm sure posts can be found in the thread somewhere but there were very many destroyed buildings/swept homes visible on high-res satellite that as far as I know, were never surveyed or at least added to DAT, mentioned in surveys etc. Even in Mayfield itself, only about half of the DIs were added. Even some structures which *apparently* may have been EF5 candidates, like some of the completely destroyed large brick churches were not officially surveyed by the NWS (as far as I am aware). No DIs for context, vehicle damage or similar were added. The recent Rolling Forks survey was a perfect example of a great survey - almost all DIS added, context, vehicle damage DIs added. However, I do realise the absolutely massive amount of path to survey in KY, but given multiple survey teams from multiple offices, I would have maybe expected more - I would never have minded waiting more for a more thorough survey.
I also found something I read on twitter this morning insightful - NWS Paducah highlighted multiple DIs for a rating of 190+mph (i.e. EF5 or at least very high end EF4), that were all kept at 190mph instead by Tim Marshall. I don't necessarily think there was much 'foul play' at NWS Paducah, I believe they did a relatively good job and they seemed not opposed to a higher rating as some surveys have been in the past. I just think there are fundamental problems with the EF scale and a pressure to be 'conservative' nowadays.