Of course it's bad science, I won't argue against that. What I call into question is the enormous selection bias at work here: modern methodological inquiries will be performed only on famous and controversial pieces that are likely to produce notable modern papers. The vast majority of bad science of those days, that happened - by accident or intuition of the experimenter - to resonate with modern science, won't be examined and remain part of the scientific history.
So "expunging" may not necessarily be intentional, but it's clearly an adequate description of what's happening.
I would be very happy if more research time was spent on the replicability crisis. I don't consider "actually this paper is garbage and we cannot trust its conclusions" to be "expunging" -- it's an improvement of science.
But in this particular case, (after I looked into it again) calling their "research" bad science is giving it too much credit. The issues are not just methodological (though they do have plenty). They outright fabricated data, came to conclusions that do not make sense given the data, and did not care at all about the health of their subjects (which was supposedly the primary point of the "study") so they didn't accurately record what effects different hypothermia treatments had (which was again the main point of the "study"). Pretending that these "studies" should be treated as anything other than pointless and cruel human torture pretending to be science is being intellectually dishonest.
A lot of science comes down to how you interpret data and you have to have a reasonable amount of faith in the academic integrity of person who did the recording and interpreting because if you can't trust them then there's no way of knowing if the data is good or not (and using it is going to lead to bad science by good scientists in the long run) -- call me a skeptic, but I personally don't trust the academic integrity of Nazi scientists who were torturing innocent people (and we know fabricated their data).
So "expunging" may not necessarily be intentional, but it's clearly an adequate description of what's happening.