I think the problems have been resolved. For my last few installations of Python, I've just used pip install for those packages, without any issues. Linux and Windows, can't comment about Mac. And the important libraries are all in Py3 now.

I haven't tried out virtual environments yet.

I've used matplotlib and pandas fairly recently because applicants used it in their code, don't remember any specific problems with that. Well, with the dependency installation that is. The applicant loading all data into RAM so that Pandas can operate on it was an issue. (The whole point was that the data was too big to reasonably load into RAM, and would constantly grow as more sensor readings come in, and the questions are such that you can easily stream it and keep a few KB of state in a dictionary to find the answers needed... alas.)

I do remember python2-only being a problem back in the day, but this was solved... hmm, maybe in 2017 somewhen? At least for the packages I used then that had py3 issues before, like sklearn, numpy, and scapy come to mind. I think it more or less coincided with Debian deciding the next release was not going to ship Python 2. Somehow that made everyone 2to3, fix a few remaining bugs, release, done. I'm too young (I'm 30) to really have done much with Python 2 so I didn't have this legacy in my own code (besides a few early files), I've ~always just tried to find and install Python 3 versions of everything.