I have been wrestling with this too. In Luminar AI you can just drop in a sky or elements and I have no idea if they are true photographs or generated images. I would say I am more comfortable calling anything made from photographs photography, even if it is composited. Where I get stumped is in manipulation of photography. I smear and change pixels, so the original image no longer remains. Is that still photography? I can also use photographs as references in Corel Painter, but then it is no longer a photograph, but a new image made with digital paint. What is it and how does that fit in? I think it is just easier to call a spade a spade, especially for competitions. I don't know if the general public need to know the nuances of everything though as that could get tedious. For instance, if a client wants an image with a sky that looks a certain way--say a hotel that wants murals for rooms, I don't see the harm in compositing it or using AI generated effects to please them (unless they specifically want a true photograph). In the end it is about selling images one way or another, right? If we are honest about a hybrid process, then I think it is ok.