I don’t see how this is different from using the pdb Python module?
You can catch exceptions and use to write functions.
I find the other differences to be minor in comparison - to be able to recover and continue execution following post-mortem debugging is amazing.