Absolutely, but a problem with legacy languages is that you can pile up a lot of cruft which gradually increases complexity.

Python today is a far more complex language to get into than when started using it some 20 years ago.

I recently began mentoring somebody building a system related to her domain. She had tried Python but gave up due to the complexities of packages, environments etc.

Which I totally understand. Despite being a professional for decades, I find the whole system around packages and environments quite complex in Python. You got the 2.x vs 3.x problem. Add to that multiple virtual environments. Add to that pip vs Conda. For things like JIT compilation you got multiple solutions.

The large number of different and often incompatilbe solutions for Python adds a lot to the complexity. At some point you will reach an inflection point where the complexity outweighs the advantages of the larger community.