Apparently, you got stuck on these pyro negatives...
...I would like to refer to the paper of M.A. Kriss, Im,age structure, The theory of the photographic process 4 th ed. Th. James ed. On p. 618, Kriss shows three pictures with the same density, but they differ in microstructure...
...In no time you are struggling with the laws of physics...
I was stuck on the shear tedium of reading stacks of pyro step tablets and the tedium of reading even greater stacks of prints of said step tablets.
I agree with you about the effects of 'microstructure' on the look of a picture. There isn't any need to invoke anything exotic - grain certainly effects the look of a picture. The HD curve obviously doesn't tell all. Staining developers, with their dye/stain cloud formation, effect the look of the grain.
The question at hand, though, is that of the gross macro effect of the stain on the HD curve. It is a question that, to the best of my knowledge, has never had a definitive answer. There is little in the literature about tanning/staining developers, but as Kodak abandoned pyro developers a long time ago it isn't surprising.
Unless the contributing effects of the 'pyro look' can be separated the issue devolves into the usual polemic about things that can't be measured. All physical things can be measured, though one may often not know how.
That's why I feel a good starting point is an instrumented approach looking just at the effect of pyro stain on the transfer of negative exposure to print tone. Start with things that can be measured without requiring the resources of Perkin Elmer.
MTF techniques may fall short when it comes to looking at micro effects in film. MTF assumes the system under consideration is linear and continuous. At the micro scale photographic images aren't analog anymore, they are a very messy digital form consisting of fuzzy brillo-pad like grains, grain clumps and, in the case of pyro, clouds of dye. MTF will, like any measurement technique, gladly throw out numbers when fed data - any data, and one has to be careful one isn't falling into the old GIGO trap.
Making proper micro scale measurement is going to require a scanning micro-densitometer. Not something most people have access to. A possible starting point might be the computer analysis of a photomicrograph of an edge - say a contact print of a razor blade.
* * *
On the macro side, a good starting point might be to look at the effect of a uniform pyro stain. Make a pyro contrast filter by bleaching the silver from a pyro negative of a blank wall, leaving uniform stain. Does the insertion of the pyro contrast filter have any effect on the HD curve of a print of a step tablet? - if not, there isn't any reason to go further with investigating the effect of pyro on macro-contrast. Maybe someone has already published the results of such an experiment?