It does make sense, if you skim through the research paper (page 11). They aren't using performance.now() or whatever the state-of-the-art in JS currently is. Their measurements include invocation of the interpreter. And parsing TS involves bigger overhead than parsing JS.
I assume (didn't read the whole paper, honestly DGAF) they don't do that with compiled languages, because there's no way the gap between compiling C and Rust or C++ is that small.
Making an assumption regarding the entirety of a social network with at least tens of thousands of users based on a single inflammatory comment (mind you, a comment that has more downvotes than upvotes)? That's a bit of a small sample to extrapolate from, don't you think? You should've probably asked ChatGPT if that's the right call.
You're supposed to use PNG for images of that sort, you fucking barbarian.