I don't buy the major argument a language is completely separate entity with its implementation, especially there is only very limited choices of implementations out there.

Other arguments feel a little too convievent to dismiss the concerns as non-issues. The author is basically saying the language can look like anything and it doesn't need to be performant, which is a pretty opinionated and not a good look for selling.

In fact we are in a programming reality that is overwhelmed with choices of languages, only expressiveness isn't enough to have people onboard.