Any time I’m thinking about a discussion of recruiting, I go back to the fact that recruiting sites are trying to do two things:
1. Their own stated goal is to try and project players to the NFL and not to college. By that very admission your criteria is much different. College football is littered with highly successful QB’s who never had requisite arm strength, height, running ability etc. that the NFL would require to be drafted.
2. They’re trying to sell subs, are unqualified to truly be scouts, and simply rely on coaches to give their “bumps” in rankings.
At the end of the day, recruiting QB’s is a tiny hit rate. What my research has shown to be most successful is a blend of data from an appropriate level of competition, track record of beating teams ranked higher than they are and/or elevate a program from previous status, and recruiting offers (this is a proxy for game film since my model cannot “watch” film and I can’t grade every HS QB in America and enter a grade).
Even then, the r^2 I’m getting on HS QB to college All-Conference at P5 level is ~.22. Much better than recruiting rankings ~.08.
My opinion is teams should be doing the research to put a process in place to identify ways to maximize their chances of hitting on an All-Conference player in college. If your process is moving your implied chances from 22% to 26%, it’s a good Expected Value move, but when you do hit on only one out of four, fans will still see it as a failure by staff and not just the inherent probabilities in play.