One point that's also omitted from your discussion is the emulsion type. For the purpose of this discussion, what you want to know is where the latent image centers are made. The most recent technology allows emulsions that make latent image on the surface of the grain, where developer is directly accessible, especially when exposure is relatively slow (which means anything but very brief flash exposure in modern standard). However, in old technology emulsions, latent image centers can be made in the grain or on the surface. Especially when the grain is exposed with very brief exposure of light (which means 1/60 or faster in some long exposure type emulsions), or high energy radiations, more latent image centers form inside the grain. Also, direct positive emulsions have latent image (intentionally created fog centers, rather) in the interior. These internal latent images are not directly accessible to your developers and therefore the sensitometric curves can differ from those obtained with developers that can access internal latent images.
I have made several different emulsions, some of which form latent image in the interior, and others near the corners of the grains outside, and I do see quite noticeable differences with the kind of factors you are looking at, including the sulfite concentration.