I didn't bring urms into the conversation nor did i ever claim to make an all encompassing statement about the relationship between LSAT score and law school performance. OP basically asked how to gauge which of his 4 schools would offer him the best chance of finishing near the top of his class. I told him IMO LSAT score is a better gauge of this than GPA because there are too many variables in undergrad GPA, whereas the LSAT is a test that puts everyone on the same level and allows for a better comparison of prospective law school success.
I sufficiently qualified my statement when I said there are exceptions and made it clear that this was my opinion based on my experience. You two are misconstruing my intentions and attacking my advice as if I was submitting it to a scholarly journal. If you both disagree with my assessment then please tell OP a better way to gauge which of the 4 schools will give him the best chance of finishing at the top. There truly is no "right" answer because there is no way to see into the future and make any guarantees. You have both made credited arguments, but they miss the point of this thread and the question that was asked.
In this situation I still believe LSAT is the best way to gauge the aptitude for doing well on law school exams because there is no better way to do it and because LSAT can be representative of numerous factors. If you disagree that LSAT score is an indicator of how well one can interpret, comprehend, and isolate key points in convoluted writing then you probably believe a high score is achievable by learning how to take the test. In which case I'm guessing you believe that since anyone can spend the time to learn the intricacies of the test, it is only an indicator of how well one can take the LSAT. If that is your opinion, I tend to agree with you on this for the most part. But in that case the LSAT doesn't just show who is good at taking the LSAT, it shows who is dedicated and studious enough to put that time in, which I think is a pretty big indicator of potential law school performance. Writing a good exam is a combination of putting in the work to learn the law, learn how to apply the law to the specifications of your professor, and put it all together and write a coherent exam under limited time constraints.
Look, I may be wrong. If so please enlighten me and the OP, keeping in mind that his question actually demands an answer. What is a better way to try and predict the potential aptitude for writing good law school exams than the LSAT? Remember though, taking shots at my opinion without offering your own isn't the sort of answer he is looking for, and telling him that there is no way to predict this is also not the answer he is looking for. He's looking for the best way to make this assessment, not a theory that can stand up to the TLS version of strict scrutiny.
BTW I didn't fail reading comp. I scored well on the LSAT and I did it through tons of consistent practice. That same discipline I used in studying for the LSAT I use everyday in law school and so far I've done well there too. I'm also not wealthy, so I didn't use any test review programs. Just old tests, all of them to be exact. I don't buy in to the opinion that economic situation is an impenetrable barrier to success on the LSAT. So unless your argument is that urms are inherently incapable of doing well on the LSAT, I don't see why it is necessary to distinguish LSAT takers in the context of OP's query.