Yes, they actually exist in the West as research reactors, since there has been no demonstration or certification of their suitability for long-term operation.

It seems to me like you're just being circular now, like you're saying "Sure, MS reactors exist, but we don't know if they are good because we haven't done the work to find out."

Going back to the original article "Why isn't differential dataflow more popular?" we could ask "Why aren't Molten Salt reactors more popular?"