Understood. I think one of the problems is that (fill in the blank: society, parents, students, guidance/admissions counselors, et al) seek to categorize and create blanket answers, whereas the correct answer is highly variable relative to the application. Boolean vs bayesian.

As a mental exercise, I'm imagining a product that would, given a number of inputs, produce a bayesian curve of the applicant's possible outcomes, and then selects the best school for that applicant. While developing this product sounds like a fun side project, I can't imagine it gaining traction, because the aforementioned groups would get upset (read: not pay) that Prestigious U is not the best choice for Johnny, rather, Mid-Atlantic Middle College is. It would be difficult to get the secondary actors in the process to pay as well, because there is value generated for them by maintaining process homeostasis.

Current thinking would suggest that I should have gone to, say, Bucknell over Penn State (again, I understand n=1 here, and I am not measuring my level of selective perception), whereas I am very happy with the course of events and think it produced the best outcome. I have to believe there is some grain of truth to my line of thinking and a potential for value generation in the decision aspect of college admissions.

The answer to the second question is pretty obviously no. There are maybe 20 private schools in the US which are genuinely worth the price over public schools. They're not just offering a comparable to better education. They're offering the children of the middle class an entree into the upper class. (Relatedly, the con which the 2nd tier private schools are running is selling parents the idea that just paying a lot for an education is a ticket up.)

All of which seemed relevant to me since the author of TFA was making sacrifices to send his girls to Emory & Stanford.