Rankings hold true far more times than they don't, you just can't get literal with them, #68 is not always better than #69. Lining up 100-300 kids in exact order is an impossible task. It also can't take into account factors no one knows like who will get injured, who will go to a system that is a good/bad fit for them, who will work hard, etc.
However, if you look at rankings at a macro level, top 25 vs top 50 vs top 100 vs top 200, now you will see very high levels of accuracy. Still not 100%, but closer. I think they need to stop saying 100 kids are a 4 star and align more to top 25, 25-75, 75-150, 151+. I think if you did 5 star, 4, 3, 2 along those lines and then just use the star level to gauge expectations, you'd see pretty high accuracy in star ratings.