I skimmed through part of this thread, though I did not read the whole thing. If I repeat anything said by others without acknowledging your post please forgive me, and feel free to point it out if you want.
Now let me make a few comments. First of all, the inverse square law is correct and applies to essentially all cases, though it may seem to be hiding in some cases.
If you consider a small fraction of an emitting or diffusely reflecting body then the intensity of the light from that body is proportional to 1/r^2. The relationship applies when the size of the little chunk of emitting (or diffusely reflecting) body under consideration is small compared to the distance from the body.
A neighboring chunk of the emitting (or diffusely reflecting) body behaves the same way, namely when you are far away from the chunk the intensity of light from that chunk is proportional to 1/r^2.
The total amount of light being emitted is independent of the distance from the light emitting body. In other words, if I draw a small sphere around the body and measure the falling on the inner surface of the sphere then the total light is independent of the distance from the emitting body, or in other words it is independent of the radius of the sphere. Let me coin a term for this. I will call it the law of conservation of radiative flux. This law of conservation of radiative flux can be thought of as the origin of the 1/r^2 law, or alternatively the the law of conservation of radiative flux can be considered a consequence of the 1/r^2 law. It really doesn't matter which you consider to be the fundamental relation and which you consider to be the derived relation. It all amounts to the same thing.
Now, I believe the basic question in the thread revolved around the question of why you don't have to change the exposure setting as you place the camera at various distances from the object, given that the light intensity falls off as 1/r^2. The reason is that the image size (in terms of area) of the object on the film is also proportional to 1/r^2. Thus, the radiative flux per unit area on the image of the object on the film plane, being the ratio of the light intensity to the image size, is independent of the distance from the object to the camera lens, and you don't have to change the exposure.
There is a special exception to the description above. It is not an exception to the laws of optics, but rather relates to the consequences of how the laws apply to certain cases. If the object is essentially a point source then something else happens. When I say "point source" I don't actucally mean that the light originates from a point object, but rather that the ideal image of the object as it is projected on the film plane is smaller than the spot size of the lens. The spot size is determined by a combination of lens aberrations and diffraction, and these cause a point source to be imaged as a small spot of finite area on the film. In the case of a point source the area of the image is independent of the intensity of the light source. However amount of light focused in the image (a fuzzy point of finite size) is proportional to 1/r^2. Thus a star gets dimmer on the film plane as the distance gets greater, and you may need to adjust the exposure to record an image of a distant star.
I have left out some parts of the explanation and a few of the fine points, but I hope there is enough there to be understandable.