[upbeat_general]

I think it highly depends on what kind of math you're expecting (and how heavily you want to rely on a CAS). If most of your work is just simplifying equations and computing numerical solutions, basically any system will do whether that's Octave, Sympy + Numpy, etc.

I haven't used Maple for a while so can't speak to it's current functionality but there's been several times I've wanted to do something in Sympy/Octave and haven't found it whereas I can almost always get Mathematica to do what I want with a quick search. I tend to rely heavily on it for some more complicated/specific symbolic operations (e.g. symbolically transforming probability distributions) and for that use case, I haven't found anything better.

I'll also say that if your use case if more numerical/programming oriented, the language used might be an important factor. I personally don't like Wolfram Language and use very few of its language features and prefer Python for anything that Mathematica isn't suited for out of the box.