The counter-example is trivial: suppose you have ten "Einsteins" in your class, but due to your dumb-ass stack ranked curve, you "cut" 4 of them. Congratulations, you just set back human progress 300 years.
For the record, UCLA was equally-idiotic when it came to grading. Technically, getting an A+ (4.3) was possible, but they were virtually NEVER given out. Why? Who the fuck knows -- I asked multiple times what I could have done better. I only ever got a handful of A+'es (and basically had straight A's).
As an aside, I wish I could nuke the entire educational system. It's corrupted by lazy and complacent professors (that clearly suck at their job: see the reproducibility crisis, p-hacking, etc.), a veritable swamp of administrators (syphoning six-figure salaries across the board), and bloated sports budgets. All while issuing unforgivable loans to record-high numbers of students.
Knowing from experience that 33% will fail is not the same thing as explicitly planning for 33% to fail. We don't have enough information to call this "stack ranking", IMO.
One of the calculus professors at my undergrad institution was reportedly asked on the first day of Calc I class whether there would be a grading curve.
He asked how many of the students were planning to be engineers. Roughly 3/4 of the hands went up. Then he said "I drive on bridges and fly in airplanes all the time. There will be no curve."
… I know, I know, it’s not exactly the same, but the logic is analogous. Hurting people so that an outcome matches some prespecified distribution is just lazy.
Of course we should have hard classes, and of course some people that try a hard thing will fail, but they should fail on their own merit.