Alan;
You are correct in understanding, but with different films in the same developer it is difficult to say that what you state is going on unless you know iodide content, silver level and etc. Otherwise the emulsion can cause the differences you see.
Kirk;
When pig gelatin was used, with an isoelectric point of about 9, no significant effect was seen in films such as described. The only examples that abound are matrix film and Kodachrome, and the Kodachrome effect is from a diffeerent source. The formation of the dyes forms a relief image.
Now, it is possible that the formation of the stain in a staining developer could form a relief image and that is pretty much what matrix developer can do, but it hardens as well so it is difficult to distinguish the relief image caused in this case by hardening and by the formation of the stain.
For those who don't know of this effect, it is well known that formation of dyes in film from external materials increase the bulk of the film where the dyes form. What is not known well is whether enough "dye" or stain forms in a tanning type developer to distinguish the relief image formed from this and from hardening. From my POV it is moot as to which causes it. The problem is that for best results it should be formed from the bottom up as it is in a matrix film situation because development takes place from the top down. A diffusion front of developer is considered to take about 15" to reach the bottom of a film, but there is also an induction period before development starts. These must both be considered before understanding what goes on.
In addition, exposure is from the top down, and the turbidity of the emulsion creates a gradient effect which is difficult to separate out from the differential effect of developer. But, over a 10' develpment time, a 15" diffusion rate is hard to assign to a 'surface developer effect'.
PE