The bits is the dynamic range potential. Changing up in post doesn't really do much. That being said, if you want to / need to do some heavy editing (like compression), you MIGHT benefit from upsampling, doing your heavy edits, then putting it back into the format you got it as. But outside of a speech or lecture, it'll likely sound more hideous than the original. I did some heavy edits of a convention lecture and that was done to the 16 bit that was that recording. The main problem was that with 16 bit, all of the soft stuff like people asking questions got lost, literally. No amount of gain adjusting could bring that content back. But every once in a while a single word or syllable would peak into the dynamic range, and you're like where did that come from. I recorded my stuff in 24 bit in the other room for a different lecture and didn't have any issues like that. But I also had my gain set a little hotter than they did.