I generally agree that kids that are truly borderline should not be takes until it's time for Plan B. I just don't think that's what is happening.
You can't look at recruiting and prospect rankings in a time vacuum. For instance, it is easy sitting here today to say Diamante Howard is not a Miami-caliber OLB or DB (after two more years of film, after he got tested at The Opening, after camps, and after his body just didn't develop). But back when he committed (in 2016), he was a 4-star OLB, 247 had him in their top 50 for the class, and he was ranked as a top 3 OLB and top 10 overall player in Florida. Nothing about that screams borderline. The problem is he did not develop the way we (and the recruiting services and a host of other schools) expected him to. The staff re-evaluated him and let the player know he should look elsewhere.
Also, I think Spring evaluations absolutely need to include a re-evaluation of the kids already committed (and there is no doubt this staff takes that aspect seriously). During the recruiting process, it is inevitable that you will get more and more information (data) regarding recruits (committed or otherwise). It only makes sense to incorporate that new information into the decision making process (regardless of whether the kid is committed or not).
My basic point is, for the most part, I don't believe the staff is taking "borderline" recruits as placeholders or anything like that. I think they take kids they project will be contributors first, then fill in their class as needed. But sometimes, information obtained later on in the process makes them rethink the player's ability to contribute, resulting in commitments. I'm not sure how you fix it entirely, other than having a blanket rule like, we won't take any commits until the start of a recruit's junior year, or something like that. But that sort of strict protocol creates has its own pitfalls, and likely results in losing out on kids who commit elsewhere (and actual develop as expected and take their commitment seriously).