Folks at "elite" universities have an extremely skewed vision of what the "college experience" looks like for the vast majority of Americans.
According to data from Drafty ( https://drafty.cs.brown.edu/csprofessors ), there are:
(100%) 377 Ivy League CS Professors
(53%) 201 Have undergraduate degrees from US institutions
(11%) 43 Have undergraduate degrees from public universities
(4%) 15 have undergraduate degrees from public universities with >50% admission rates.
~16,000 students will enroll in an Ivy League university this Fall.
~1,000,000 students will enroll in public universities with >50% admission rates.
Let's assume that undergrad college admissions are unbelievably good, and that their ranking of applicants explains 80% of the true "intellectual potential" of the application pool (I do not think anyone would estimate this value to be higher).
About 2% of the US population has a PhD. What percentage of the PhD-having population would you expect to be from each bin? Assume that admission ratings and true "intellectual potential" are jointly normally distributed (almost certainly true at this scale), and that there relationship is linear with correlation = sqrt(% of variance explained). **Under these conditions, you'd expect the PhD population to split 50/50 between these two groups!**
Having had the privilege of interacting with undergrads at schools from each category I listed above, I can confidently say that folks at "elite" universities drastically underestimate the top-K graduates from large public universities, and that folks at large public universities drastically overestimate the bottom-K graduates from "elite" universities.