Its a proxy for having practiced CS questions, not being able to handle the complexities of a modern software project. They basically want to know whether you went to a reputable college, or have a similar skillset to those who do.

Well, maybe I should have said it was an attempt at a proxy for general intelligence? For what it's worth, I agree with you. When people start studying for interview questions, then something is probably wrong..

I have a few friends from large universities (waterloo, stanford) that have left their current job to specifically study for a new job.

My assumption is that there are a lot more people who are studying for these interviews than you think, which is why the proxy is failing (or producing homogeneous results)

That's true.

On the other hand, it's okay to target sufficient but not necessary. False negatives can be fine, so long as you are finding enough true positives.

Even if you are systemically missing out on a sub-pool who are good, it's not necessarily a bad thing - particularly if you are realistic about its limitations. To my mind the real problem isn't that you might miss good people (you will) but that if your filter is too constrained the good people you find might be too homogeneous.

You can also cram for IQ tests the only ones with validly are those taken cold

How much cramming would it take an average person to score 190 on a quality IQ test?

There are no quality IQ tests that go up to 190. If a test purports to be valid up to that level it is very strong evidence that its authors are frauds or don’t understand statistics or both.

Depends on how far from 190 you start.

Assume 100 for "average"