Something to note here is that most (if not all) of the "medical data" acquired by Axis experiments is useless: a lot of it is on the order of "if we make someone really cold they die". The methodology was, unsurprisingly, generally biased, non-reproducible, and often cruel for the sake of it, rather than unethical out of necessity.
IMO there's a nice parallel between useless evidence from bad experiments, and useless business practices from unethical companies. If you want to take the lessons but leave the bad stuff, often you'll find there's nothing left.