The reason for the limited rotation is that it is really only a visualization trick - the coordinate system used by QuPath remains the same, with (0,0) in the top left corner of the unrotated image.
A common reason to want to rotate the image 180 degrees is to ‘correct’ an image (e.g. a tissue microarray) that is ‘upside down’.
If this is attempted just with the rotation slider the results and behavior could be surprising because internally all of the analysis and annotation code doesn’t know (or care) anything about what rotation was applied for visualization. QuPath would treat the bottom right of the image as (0,0) and all numbering would likely be the opposite of what is expected.
The purpose of the hack was to effectively rotate the image fully, so that the image behaved completely as if it had been scanned the other way around.
I’ve written the code to implement the hack more cleanly and offer 90, 180 and 270 degree rotations. I plan to incorporate it into the interface but haven’t got around to it yet.
Are you asking because of a rotated tissue microarray, or is there another reason why this would be important?