employing the scientific method

Really? They have control groups? Blind and A/B testing? Hypothesis that they set out to reject?

I'm sure they have methods but are they scientific?

Depends. A proper computer science course is basically math with machines. At the highest level, it may have zero programming at all, and the machines in question are entirely abstract.

Software Engineering is, well, engineering (setting aside the whole debate on what makes a "real" engineer).

It used to be that universities crammed both under "computer science", and you had to look at the curriculum to figure out which one they were actually teaching. They tend to separate the two more clearly these days. Neither is really "science" in the strictest sense, but the term stuck now.

“Real” scientists try to put a spin on it akin to “You can’t properly hypothesise, reason or make predictions about anything based on a sample size of ~200 countries that are totally outside of your control and are very different from each other”. Few more arguments get thrown into a pot.

Doesn’t stop political scientists from mostly accurately describing things, so no harm is done here. The harm lies within pushing that opinion on general public, highlighting the that “proper” scientists don’t see any value in social “sciences”, hence contributing to public ignorance about societal problems.

And with how lousy political views of “rational”, “logical”, “critically thinking” people in STEM sometimes are, it’s awfully ironic.

Speaking as a disgruntled Russian STEM scientist who is horrified how willingly some of his collages ate Putin’s reasons for actions both against Ukraine and within Russia, including against fellow scientists (WTF, where’s professional solidarity?!).