The methodology used in this paper was extremely bad. There is no reason there should be any disparity between a TypeScript and JavaScript benchmark.

Unfortunately it has continued to make the rounds for about a decade now and gets re-posted every year or so.