I've always had two threads in my work with photographs, from back when I shot film, to now with digital.
Thread one - fairly straightforward capture of a moment or a vision. Some adjustments of color balance, saturation, brightness, contrast, including brightness and contrast at a local level, all with the focus of making the photograph look like what I saw. (And since my brain does not auto white balance like most peoples' do, I've often taken flack for saturation and color manipulation when all I did was shoot daylight film with no adjusting filters, or today, keeping my camera and raw processor set to daylight.) I wouldn't have airbrushed out anything then, and won't clone it out now - cropping is fair game, as is diffusion filters, vignetting, and the like.
Thread two - fairly extreme manipulation of captured images to achieve a statement or vision beyond what was photographed. Trying to elaborate on what something FELT like when I saw it. In the old days, I did Kodalith, photo silk screens, cross processing, solarization sometimes in multiple iterations, hand coloring, and sometimes many of them at once. I put things in more than removed them, but today I may well clone things out as part of the process. I've even written code to transform digital images.
When I sell the first, I call them photographs. When I sell the second, I call them manipulated photographs, which is somewhat unsatisfactory to me still, but I think at least gives the viewer fair warning that they're going to see inside my head, not so much what my eyes saw.
I've been aware for a long time of the huge gulf between the two, but feel like if even once I cloned out a wire or a random foot that made its way into an image that I'd find myself sliding down a slippery slope. I've done so a few times - removing the red lights of the radio tower that juts up right next to the cathedral in San Miguel de Allende - and I always find myself really uncomfortable when I do it, and feel compelled to tell people viewing the image that I did it. It's an odd quirk.
But, as a long time admirer of Jerry Uelsmann, what would you call what he does? Yeah, it starts with photographs, but in the end it's not captured light so much as output light. John Paul Caponigro is another benchmark of mine - are they photographs? In both cases the capture is less important than the processing and print.
I'm acutely uncomfortable with the idea of tweaking mountain height, moving trees, things like that, where the subject area is "landscape photography." It wouldn't change my appreciation of the result, I think, but I wouldn't be thinking of it as a photograph. As a print? Yeah, but there's a modifier needed before the word "print." Don't know what that modifier should be.