These "services" have been junk for years. Remember they told us that Mitch Mustain and Brock Berlin were all-world QBs. Remember the ultra-high ratings for Anthony Morelli and Tyler Palko. All these guys were fine players, but never lived up to the massive hype. Meanwhile they totally missed on Aaron Rodgers, Ben Roethlisberger, and Tom Brady. That's a pretty pathetic track record. They have absolutely zero credibility in talent projection.
I'm not sure your list of QBs says exactly what you think it should:
1) Big Ben is possible the most overused false example of how the recruiting sites get things wrong. When he was in HS he had a QB a year ahead of him who also happened to be the coach's son. Because of nepotism he didn't play QB until he was a senior. In fact he was recruited by schools as a TE because he was playing WR/TE in HS until his senior year. I guess you could argue the recruiting services could have noticed him at this point but none of the major conference coaches did either as he ended up in the MAC.
2) Rodgers might be just as bad an example. Do you know how many D1 offers he had coming out of HS? The answer is ZERO. Do you know why? Well he is 3 inches taller today than he was as a recruit. Starting in his senior year and through his time at a JUCO he had a (late) growth spurt that also gave him the frame to put on a ton of weight.
3) It seems the myth of Tom Brady being a nobody before the Pats continues to grow. First of all, Rivals, Scout, etc. didn't exist when he was being recruited back in 1994 (he was in the 95 signing class). What did exist was Super Prep which rated him as the #63 player in the West and a top 5 QB in that region. Today that would probably be the equivalent of low 4* or high 3* at worst. To be fair being a low 4* doesn't sound like a guy who projects to be one of the best QBs in NFL history but it's not exactly like he was some 2* under the radar guy who no one ever heard of.
4) I'm not sure what ratings you are using on Palko but I looked up his Rivals ranking and they had him as the #8 Dual-Threat QB and outside the Rivals 250. He was a multi-year D1 starter and spent a few years in the NFL. That sounds about right for a guy with his ranking. BTW, Vince Young was the #1 dual threat that year...think they nailed that one.
5) Berlin is an interesting one because UF also signed another top 100 overall player and top 5 QB in that class. So something had to give and it did. The “other” highly hyped QB won out, became a 1st team AA, Heisman finalist, 1st round pick and NFL starter. Berlin eventually transferred, became a multi-year starter at another school and after going through multiple OCs in college still found his way onto a NFL roster. Maybe he didn’t live up to all the hype but he wasn’t a total bust and the other highly ranked kid he went up against at UF surely lived up to ranking.
6) Mustain is another matter of circumstances guy. He started his true freshman year and then made one of the strangest transfer decisions ever by going to USC where they had a stable of other highly ranked young QBs already there. If he had just stayed put then he likely ends up a 4 year starter and with a legacy closer to the hype.
7) Ok, I’ll give you Morelli but even that has a story behind it. He and Henne attended a camp/combine together (I think it was a Nike camp) in which the QBs were allowed to bring their own ball(s). Henne did not know this and ended up using one of the new and slick balls provided by the camp while Morelli used his worn in one fit to his hand. Henne came in with all the hype (not that Morelli was an unknown but was not considered in his league) but because of the circumstances Morelli out performed Henne. So what were the analysts going to do…drop Henne, ignore the performance or elevate Morelli? They went with the last option and were wrong. Now there were rumblings that Morelli’s FB IQ was…well to be nice…not high…but I think the analysts wanted to avoid what might come off as calling a kid stupid so I think they ignored it. Since, I’ve seen them make more references to things like that so maybe they learned their lesson?
So I guess my point is that sometimes anecdotal examples don’t tell the whole story. I personally don’t think the rankings are an end all or even close to perfect. I think on a macro level they do a pretty good job of predicting…5 stars (overall) work out significantly more than 4 stars, 4 stars significantly more than 3, etc…and teams that tend to recruit more higher ranked classes tend to win more than those that don’t. But on a micro level one shouldn’t fool themselves into thinking it is gospel because it isn’t. However, it is not always because of a misevaluation (though it can be) and sometimes there is more behind the story.