I'm pretty sure that 80% if what we learned from the Nazi/Imperial Japan super unethical experiments was "what can a psychotic doctor justify in order to have an excuse to torture people to death."

Maybe 20% was arguably useful, and most of that could have been researched ethically with other methods.

The potential value to the Americans of Japanese-provided data, encompassing human research subjects, delivery system theories, and successful field trials, was immense. However, historian Sheldon H. Harris concluded that the Japanese data failed to meet American standards, suggesting instead that the findings from the unit were of minor importance at best. Harris characterized the research results from the Japanese camp as disappointing, concurring with the assessment of Murray Sanders, who characterized the experiments as "crude" and "ineffective."

https://en.m.wikipedia.org/wiki/Unit_731

To back up your point that the research gained by unit 731 was useless.

“People die if you kill them”

This one was making a child with an HIV-positive parent resistant to HIV, so it's a bit better than 731 torture.