>Assuming a grad income of $80k is an insane starting assumption.
It doesn't really matter if you consider it to be insane. The studies on this stuff always compare averages to averages, and average college grads do better in the long run no matter how optimistically you cook the books to make the inverse seem likely.
Here is my made up guess: The average university graduate will do better without degree than the current non-grad averages.
You are making the assumption that it all boils down to the degree (the difference in income). This can't be possible (maybe a majority of it, but not all of it). There are other factors (like being from a middle-class, higher IQ, etc.) that selects for going to Uni. and this has effect later on income.
> average college grads do better in the long run no matter how optimistically you cook the books to make the inverse seem likely
This is far from true when you consider the selection bias of who goes to college, and the sheepskin effect.