Thanks for reading!
You'd be surprised how many people have commented on this (without reading) about other assumptions that I call out, but you've spotted the main issue with the PSNProfiles Bias. :)
I actually spent months scraping user data from the PSN instead, and while there is still some bias in that it is far less than PSNProfiles (people can opt out of me scraping their profile on PSN).
From there I was able to get a better representation. While the numbers aren't quite 25%, they are still high (at what % would be an acceptable level in your mind for this statement to not be shocking :P). There are definitely people still sitting on PS5s whether for vanity or due to scalping.