I think it is less an issue of STEM vs humanities and more of an issue of empiricism vs idealism/rationalism. Biologists or physicians might disparage the human

I think it is less an issue of STEM vs humanities and more of an issue of empiricism vs idealism/rationalism. Biologists or physicians might disparage the humanities (and often they definitely do), but by training they are wary of any great theory of anything extracted from three anecdotes. Whilst in this circles there are people who don't deal with empirical evidence at all (programmers) or who do so in fields where you can safely generalize from a single piece of properly obtained evidence, like physicists. Another problem of non-empirical science, or sciences that happen only in the lab, is that nobody prepares you to deal with evidences of uncertain, or definitely poor, quality. A biologist knows that a lot of samples from the field are taken by humans and there is some margin of error. Or maybe they are contaminated. Or maybe conserved poorly. And so on. So they can perfectly understand historians who say how much of our "data" about pre-1900 societies are poorly collected, guesses themselves, or manipulated for political aims, or simply modern inferences based on archeological data but subject to interpretation. Instead a programmer or a physicist looks at a graph with "number of laws, 1000-1900" and "GDP per capita,1000-1900" and think that the quantities represented are actual, objective data and that there are not 18 different and equally sensible estimations that would make the graph look completely different. Combining these two factors, you get all these posts that supposedly explain the iron laws of every society ever with a bit of deductive logic, some "common sense" arguments the reader can relate to, and three graphs that show correlation and not causation between variables nobody actually measured