Several here have stated several times that the recruiting rankings can be misleading because they value quantity just as much as quality. Of course, until NLI day, recruiting rankings are constantly in flux. So, beyond the idea that early momentum helps you to finish out your class stronger, the ranking that you have at this point in time means very little.
That being said, IMPO I think that all the recruiting services should re-work their team rankings based on average star ranking instead of total points = stars x quantity. The reason is that you can fairly safely assume that all teams by NLI day will have filled their roster with the max number of recruits. Very very few schools are going to purposely not fill their class. So, other than scholarship sanctions or other oddities, you can assume that every NLI class is full. If every class is full, then you can assume that every school brings in the same number of recruits over a 4-5 year span (yes, I understand that transfers, dropouts, suspension, early entry for the draft, and non-qualifiers all affect attrition and higher attrition = more recruits brought in - so not sure how you account for that in recruiting rankings). So, if you go back to the basic assumption that every school has the same number of recruits on scholarship over a 4-5 year span, then the numbers are equal and "quantity" part of the team recruiting ranking becomes relatively meaningless. So, a team's recruiting ranking should be based on average star ranking, and a rolling 4 or 5-year average star ranking is even more meaningful. Perhaps a rolling average could weight the class from 4 years ago more heavily because they have had more time to develop and become starters whereas the true freshman that just signed likely will have little impact that year. I know that Sam McKewon annually does a story on how productive a particular NU recruiting class was based on how many starts they attained, starter and 2-deep status, attrition/busts, etc. That is really an analysis of how good a class turned out to be instead of a ranking of the recruit's potential. But, I could see a rolling team recruiting rank weight a class from 4 years ago more heavily but then perhaps to gauge the potential of that class (which is what a recruiting ranking is) you only count the members from that class that are still on the team when the rolling average ranking is calculated each year. I think those means to rank recruiting classes would be much more meaningful and accurate than the current system that the services use of ranking = total points = quantity x quality.
Yes, a team can recruit up to 25 players per cycle. But think about this, if a team like Clemson actually had zero attrition for 5 years and managed to redshirt each entire class, they could only recruit 17 players a year or they would go over the 85 limit. Yet, they would still be full. If they only recruited 17 5* players every year, then they would have an entire team of 5* players. Yet, if Alabama recruited 17 5* players and 8 4* players every year, but they had a 50% attrition rate every year, Alabama's roster would not be full and they would have a mix of 5* and 4* talent. Yet, according to the recruiting rankings, Alabama would be ranked ahead of Clemson every year.
Yes, this is an extreme situation, but in reality there are less extreme versions of this happening all the time. That is why average star rating and rolling weighted average star rating of recruits still on a team are more meaningful indicators of the potential talent on any given team. That's what recruiting rankings are really trying to show, but are missing the goal. The rest is coaching, scheme, luck, development, projection skills, SOS, etc.