[dnautics]

You're right, that a lot of grad students whittle away at time doing stupid things. But your logical arrow is pointing the wrong way. Except in some fields (NMR: pop your sample in the machine and go surfing for seven days) I haven't seen a successful student who didn't spend a lot of hours working on things. A lot of grad students wasting time and being unsuccessful doesn't mean that being successful doesn't require not wasting time.

An oversupply of labor should have made the time problem evaporate.

Exactly. It didn't. That suggests the time problem is resistant to labor parallelization.

I think that oversupply is already not very high quality at this point due to the fact that you are treated better and paid more in private industries.

You're correct that the oversupply is not very high quality. But it's not because people are treated better in private industry. How many grad students quit mid-gradschool because "private industry pays better". How many postdocs quit mid-postdoc because "private industry pays better"? The market won't correct for this so easily because of several reasons. 1) the government incentivises the hiring of grad students and postdocs and channeling labor into that market 2) there is a narrative disconnect where professors don't give the 'real story' to people intending to pursue an academic career, partly because they are disincentivised to do so (see 1) and partly because there's selection bias (they got lucky, but don't realize it, and think that anyone can and should do what they can do) and 3) there's an irrational prestige drive that channels labor into this pool. How much propaganda do you see that science is awesome! And science is cool! And I fucking love science! And how much propaganda do you see that the life of a scientist is miserable, shitty, and doesn't pay well?