I think it is true that Yale and others at least often look at relative performance in your HS when forming an internal academic rating.
I don’t think this observation is inconsistent with the idea that such a rating mostly plays a “good enough” role in their admissions decisions.
I think the complexity comes from the massive variety in school curriculums, grading systems, grading norms, and so on. And then there are many other factors they also may consider in forming an internal academic rating.
So guessing what sort of relative GPA you would need to get a good enough internal academic rating at Yale is often going to be tremendously difficult. But if your HS collects data like that, and has a large enough sample, you may get some clues from that data.
As usual, I am not sure there is anyone arguing otherwise. Like, I have frequently referenced the SCOIR data available to me for our HS, and have specifically discussed what it seemed to be suggesting about the competitive ranges for different colleges.
Obviously everyone understands this is only directly indicative for that HS, but I don’t recall anyone suggesting this was pointless for applicants from that HS. Quite the opposite, whenever someone is asking about chances, and it is apparent they go to a similar HS, usually multiple posters will suggest they should really be working with their HS college counselors, who have this sort of data and other information they can use.
By the way, I didn’t do a formal study, but I would say what I saw in our SCOIR data was broadly consistent with the good enough model.
Often there would be a more or less detectable range where inside the box (high enough GPA and test score) the frequency of admissions was decent. But it was not the case that all the highest people in the box were accepted. The data really doesn’t have enough depth to be sure what happens in the box, but it definitely was not inconsistent with all the applicants in the box having roughly the same chances.
Then outside the box, acceptances were rare, rare enough to be mostly just hooked. Not good enough.
The details were the more interesting part. Some colleges seemed to have tighter boxes and harder boundaries. Others were much more fuzzy. I think the most natural interpretation of that was some colleges had a stricter sense of good enough than others, at least for our kids.
Anyway, my point again is just that I don’t think there are many, if any, people here arguing that relative performance in your HS is never a factor in highly selective college admissions.