Thanks everyone for responding so far. Interesting points. Where I'm coming from on this - we have an implicit (or explicit) assumption in any sort of sensitometric study of a film/developer combination, whether it be ISO, Zone System etc. that whatever we measure as base fog density can be subtracted from all exposure densities to determine net density (or image density). In other words, suppose you plot a characteristic curve for a film. Base fog density is assumed to be constant all the way across. In a Zone System-style test, we are usually told explicitly to subtract this base fog density from all exposure densities. In the ISO triangle, it is not an explicit procedure because we are concerned with low densities above base fog (speed point) and an increase of 0.8 above m at an exposure 1.3 log H higher. But there is still an implicit assumption the delta D of 0.8 is "usable" or "imagewise" (pick your term) density, meaning fog is assumed to be constant.
Said another way, we know base fog (zero exposure) for a given film type can sometimes vary with age, hold time, developer and development CI. Can it also vary with exposure?
Example - suppose you do a sensitometric study of FP4 developer in ID-11. You measure fb+fog density to be 0.30. Choose any other exposure value on the curve, say one that produced a gross density of 1.50. Is the base+fog density still 0.30 at that gross density? Are we correct in assuming net density at that point is 1.20? Then change the developer to something with different developing agents and/or a stronger alkali and repeat. What then?
The emulsion could be a variable, and the developer could be also. We know certain developing agents and/or formulas have tendencies to "discriminate" better than others when it comes to exposed vs unexposed silver halide grains. Is it possible in areas of high exposure where lots of reduction is occuring, more unexposed grains might be "infectiously" developed?
Granted, even if this does happen, the differences are probably very small, and I don't know how you could measure it anyway. So yes this is probably unimportant in the grand scheme of things, but I'm still curious. PE once told me (if I'm remembering this correctly) that when they tested films etc. at Kodak they always plotted gross density.