I revisited this a little bit yesterday, experimenting with my Fallkniven DC4's ceramic side, which had a few glazed spots on it from use. Used a piece of fresh ~320-grit SiC paper I had at hand, over glass; the end result was the ceramic being somewhat more glazed & polished than it started out. As mentioned, if sandpaper is to work well at all, it's going to have to be much, much coarser than the desired target finish, something like sub-120 or so, for a finish like a medium Spyderco.
When mentioning how little swarf is generated, especially in trying medium or finer grits of SiC paper, I mean not even enough swarf to darken a clean microfiber towel after wiping down the stone with Windex or oil. If it can't even generate enough swarf to do that, it's a clue the abrasive isn't cutting very well as used in that form. Additionally, the paper backing of the sandpaper cushions the abrasive enough that it's even more diminished in it's ability to cut the alumina, which just compounds the tendency to overpolish, even when using relatively coarse grits. Same effect can even be seen in sharpening knives on sandpaper, leaving an effectively finer finish than if sharpening the same blade on a stone of the same rated grit.