How exactly does it help rip sprites? As I understand it, it requires me to have two screenshots with the same sprite frame but different backgrounds, and then align them so the sprites line up, so the "diff" shows just the sprite. I presume the main reason for this is to deal with the "black on black problem" when you disable background layers. It may also mean that in cases where "sprites" are not really sprites, eg. they're "burned" into background layers, or it's just a PC game without sprite hardware, there's a possibility to isolate them automatically. But how does it make it any easier than just pixel-pushing? If I can make my sprite appear on different backgrounds then I can just choose one that doesn't have black, and flood-fill may way to isolation. That, or harass emulator authors to give non-black backgrounds.
Perhaps what I'm saying is: can you give an example?