These things I don't usually look up for Python, because Python is clearly not a very functional language and doesn't have the immutable data structures at all, and doesn't even feature TCO, so basically Python is mostly lost, when it comes to doing FP, and I accept, that it is not really that kind of language. I only try to use FP knowledge to avoid common pitfalls, when coding in Python. But from Lua I would expect better.
not sure which means more, but the second is definitely critical.
While it was already in widespread use, Python really took off in the scientific/research community, thanks first to numpy and then to all the other libraries built on top of it up to the current crop of ML/AI libraries.
Those people used to write a lot code in Fortran, Matlab and R (depending on their research area) which are all 1-indexed.
Also Lua is kind of frozen in version 5.1, if one cares about performace, while even if CPython only recently got JIT love, there are several other alternatives.