Using a crappy acapella to create a good quality acapella

I’m very new to music production, but I do know a bit about audio. Anyway, so I noticed that many acapella tracks are terrible quality. One thing I see a lot is versions of the same track at different bit rates, but when I check the spectrogram, they’re clearly bad transcodes.

So I was wondering, say I have a crappy compressed acapella track, along with the normal track as a perfect FLAC, is it possible to use the acapella inverted to remove the vocals from the original track as much as possible, then invert what’s left and add it back to the original track, getting a better quality acapella track in the end? Is this nonsense or is it common practice?