It doesn't on Rivals. They only consider the top 20 recruits for team rankings, and most teams sign 20.
It's a little less clear on 247. It appears it's weighted so once the team is beyond somewhere between 20-22 recruits, it also matters very little to the overall ranking.
That still matters though, players 20-28 fill out a roster.
Here is my point (for this hypothetical, let's assume that recruiting rankings are 100% accurate in predicting talent):
Team #1:
Year 1 - 25 recruits (20- 3 star & 5 - 2 star)
Team #2:
Year 1 - 20 recruits (19 - 3 star & 1 2 star)
Team #1 will likely have a better team recruiting ranking for that year even though it has a worse star average.
Year #2 - Team #1 has attrition, loses people and signs 28 people
Team #1:
28 recruits - (20 - 3 star; and 8 - 2 star)
Team #2:
21 recruits - (19 - 3 star; and 2 - 2 star)
Again, Team #1 has a better recruiting ranking for Year #2. However, because of attrition, they actually have a worse star ranking and less experience. A coach who continually has attrition issues (see the context of what you're quoting) will always have this false advantage in terms of recruiting rankings.
I'm not saying that is what happened with PJ, I pointed out that I had never heard that before. I'm saying if a coach has continuous attrition issues, he could easily have 5 straight recruiting classes that ALL have higher team recruiting rankings than another team and that other team has higher rated and more experienced recruits.