I guess, but not really my point. All the projects that go through massive rewrites to port into another language always seem to attribute the gains they inevitably find to the new language and not to the fact that they completely restructured the code in full view of what a completed version already looked like. Very much a pilot sytem result [1].

I know there are plenty of rewritten projects that truly benefited from a language better fitted to the domain, but rewriting itself will usually produce better code. The argument against refactoring would be that the better code isn't often worth it when you're not changing languages.

[1] https://en.m.wikipedia.org/wiki/The_Mythical_Man-Month#The_p...

Edit: I submitted accidentally.