This made me think of something I learned in economics class last semester, when she discussed a study that her and a few other professors conducted.
https://www.msu.edu/~dickertc/301f06/SAT.pdf
It's quite interesting (I am an economics major though, haha) and it discusses the economics of the test-optional policy, in relation to theory (of voluntary disclosure) and in practice.
Voluntary disclosure is this - people have incentive to reveal their SAT scores despite the fact that their SAT scores may not be up to par, because people don't want schools to assume that their SAT scores are worse than they actually are. Therefore, it is not necessary to make test scores mandatory for colleges, because students realize this. Say, School X (with a test-optional policy) has an average SAT of 2000. Students that submit their score may have an SAT lower than that - 1800 for example. But they still choose to submit their SAT score, because they assume one of several things; 1, their SAT score is representative of their ability as a student; 2, their SAT improves upon the rest of their application; or 3, revealing their SAT score would prevent the college from assuming that the student was a worse student than they actually are.