--snip--
I don't see the point of using a "python accelerator" written in C/C++ when the same tools could have directly been written in C++.
Besides, if you got a Cpp program, you can make language ports to it to Node, Rust, ... as well as Python.
If you already write your tool/script under Python, those "python accelerator" could be very useful if you don't have sufficient programming skill on faster programming language or simply don't want to port it manually. Those accelerator usually only require no to small change anyway.
It's simply a fact that Python is slow; it's an interpreted language. 10 to 15 minutes may still be acceptable in this case, but what if you need to analyze 1 million private keys?
python can handle 1 million private keys.
Probably when the library you use has C/C++ backend or you use some kind of accelerator.