There may be a very simple reason for why Law Schools are giving a lot of weight to LSAT and GPA, which I have not seen addressed here. I know the current trend is to question the effectiveness of standardized tests in predicting future academic success in college, but from what I have read, it looks like a combination of GPA and standardized scores is better at predicting academic success at college than either one of these alone. So if you are concerned with recruiting the best class purely from an academic perspective, you would definitely look at these two factors closely.
And maybe they are even bigger predictor in Grad schools specially professional schools and maybe they directly correlate to the success at clearing the bar exams. Maybe both USNews and the law schools think are on the same page on this, hence both give those metrics a lot of importance because it generally predicts the strength of the incoming class and predicts the percentage of students who will easily clear the bar in their first attempt.
So just because USNews and the Law schools give a lot of weight to the same factors, does not mean USNews is dictating the preference. It may just be that both agree that these metrics are important
Having said that, One other thing that could be hurting the 3.6 GPA could be “relative comparison”. If the recommendation letter from a faculty member at one school rates the 3.6 as “average” compared to his/her peers but the recommendation letter from another faculty member at a different school rates the 3.8 as “superior” compared to his/her peers that could make a difference, even though, if they came together in the same class, it is very possible that the 3.6 applicant could run circles around the 3.8 applicant.
This may happen if the the peer group against which the first applicant is being compared is exceptionally talented academically while the peer group against which the second applicant is being evaluated is talented but not at the same level. They may both be from good schools, but given the way the undergraduate college shapes their incoming classes, there might be small but significant differences in talent, specially if one school is more skewed towards “Holistic” admission and one is more skewed towards “academic” profiles. The latter school may end up with very pointy academically talented kids which may make it harder for any single student to stand out, compared to the the academically oriented student at the former school. For e.g, I do think that it is harder for a student at Caltech to stand out academically in his/her peer group compared to a student at Brown or Harvard because of how these schools pick their undergraduate classes. It may thus make sense if you are interested in law school to go to a school which recruits its undergrad class from a wide range of academic profiles, because it is more likely that professors at these schools will see the academically talented kids in different light compared to the other kids. (I will readily admit, this may be my personal bias)
A cursory look at SAT scores does seem to imply that Chicago may be getting kids who are more academically talented than say kids that go to Harvard (on average), not because there are’nt academically talented kids are Harvard, but because they may be a proportionally smaller percentage than at Chicago because of athletic and other holistic forms of recruiting at Harvard. So a pointy headed academically oriented kid will stand out to his/her profs at Harvard than this kid would at Chicago and may get stellar recommendations.
Both are being assessed “relative” to their local populations and that may make a big difference. The only standardized metric then is the LSAT and if they are about the same, the 3.8 with better recommendation letters may be given the edge, not because of the GPA but because of his/her relative position in class. Without seeing the recommendation letters, it would be hard to figure out that it was not the GPA but the recommendation letter that finally tilted the scale.