There has been so much in conflicting opinions bantered about that I decided to conduct tests of what the staining effects amounted to with PMK developer. These tests were verified by using an Xrite 310 densitometer.
First of all we must determine what aspect of stain that we are speaking of. If we are talking about proportional stain then the tests would support that development in PMK and other Pyro based developers does indeed depend on silver density on the negative. This silver density is more a function of exposure and development then it is of the matter of whether a film is more or less "silver rich" (whatever that means.) I did not find that the effects of the proportional staining was dependent on film manufacture or film speed so much as it was a condition of the density of the negative. My tests indicate that the effects of proportional staining typically add nearly .28 (when read through the blue channel) to a negative as compared to a visual density reading of 1.20. The effects on UV transmission are still greater. The effects of this proportional staining are not readily discernable to my eye. I have determined the effects of proportional staining on not only TriX, but also FP4, Bergger BPF 200, and JandC Classic 200. The effects of proportional staining on various films is typical very near the same at a given density with a given developer.
If we are speaking of general stain as compared to proportional stain. (of which most people seem to be speaking) the staining effect is discernable to the human eye and is not proportional (in and of itself) to silver density of the negative. This would tend to indicate that the presence or degree of this general stain is not as dependent on film type (it's relative speed etc.) as it is upon film base material and of the carrier in which the silver emulsion is suspended.
This is born out with testing experience with 35 mm film in comparison to sheet film. The base material of 35 mm exhibits a density reading of .15 film base and fog as compared to .04 with sheet film in the case of TriX developed in the same developer. The unavailability of proprietary information from Kodak as to the base material characteristics does not enable me to evaluate whether the cause of the general stain effects attributed to PMK are those of the film base material or of the emulsion carrier material. It quite apparently must not be due to the amount of silver or of the film speed. The reason is that the general stain exists in regions of low silver density (TriX sheet film) and not to the same extent on the Tri X film in 35 mm format. It is interesting that this film (TriX) is not a slow speed emulsion and that the film manufacturer is the same in this case.
In the use of pyro developer formula ABC we can reach a point in development at which fog levels increase to the point that the effect is one of effectively inhibiting the contrast within the negative. This fog development does not occur as rapidly with Pyrocat as it does with ABC. In the case of PMK or for that matter any developer that exhibits the tendency to impart general stain the effects are akin to the formation of general fog as I have indicated above. The formation of general fog and general stain have the same generalized effect that being the inhibition of the full potential of proportional staining.