Kenpom is the one everyone knows because it gets mentioned by far the most on TV. Torvik and Evan Miya are the other two that chronically online basketball fans know about.
The Boneyard has a subsection of fans who were online before advanced metrics were well known or existed at all, and they probably don't care much about any of that stuff and care more about traditional media (i.e. TV and reporters). I guess what I'm saying is if you were in your 20s when you found the Boneyard 25 years ago, you probably don't give a crap about who Torvik is lol.
Anyway, these models are going to be smoothing out year over year improvement, etc. Reed was statistically a pretty mid player on a bad team and if you assume he's going to get like 5% better then UConn is not a top 5 team.
A real human might feel confident saying Reed can be one of the best big men in the Big East after being coached by Hurley, but if he transferred to Georgetown we would not be projecting him as such. I doubt this model has any reliable way of capturing that sentiment/reality.