I believe I've written about this before. Or maybe it's at my website. But it's fairly simple, so I'll repeat it here. The pouch dishing tools I've bought (and sold) create a dish in the pouch that is 1/8" deep in the center regardless of the diameter of the pouch. I don't know if that's a standard depth or not, but it will certainly work for all note valve pouches in any player piano -and here's why.

As Art Reblitz noted, the Standard Q&A suggests a pouch-to-button clearance of 1/16" (0.625"). Doing a little bit of math, if the center of the pouch has a depth of 1/8" (0.125"), it will protrude out of the pouch well by 1/8" when inflated -provided it is properly glued at the edges. That's a total movement of 1/4" (0.250"). The maximum valve gap that I've ever measured was 0.055" in a plastic Aeolian block valve. The median valve gap is 0.032"-0.038", with some as little as 0.028" and others as great as 0.042". Here's my point. If the valve button has a shape that is identical to the shape of the dish in the pouch and there is a 1/16" gap between the pouch and the button, even when the valve is pushed up as much as 0.055", the level of the pouch will not be higher than the level of the surface onto which the pouch is glued.

As Art also noted, the pouch has it's greatest amount of power from time Zero. In other words, the further the pouch travels before it contacts the button, the less power it has to move the valve. This would seem to indicate that the optimum pouch-to-button clearance would be 'next-to-nothing'. However, we also have to take into account changing environmental conditions. Leather, like all other materials, expands and contracts with changes in temperature and humidity. Leather also shrinks as it ages -or dries out. So, in order to accommodate these changes, it was found that 1/16" was a good minimum pouch-to-button clearance. Also, as most likely has been noticed by everyone who has remove a pouch board, the shape of the pouches, when the board is first removed, is anything but the shape of a dish. In fact, the pouches always seem to appear as though they have conformed to the shape of the button. This, in fact, is exactly what happens with less than a thousand on-off repetitions. For reasons which I cannot explain, the area directly below the button conforms perfectly to the shape of the button and the remaining area of the pouch 'puffs up' in a wave shape as high as possible. It's interesting to note that the height of this 'puffed up wave' is determined (or limited) by the depth of the button in the pouch well and the initial pouch-to-button clearance.

Now, even though I've never performed tests to prove this point, it's my sincere belief that the power curve of a pouch is not linear. In fact, I believe that the power drops off significantly after the pouch reaches 50% of its total travel. If such were not the case, and taking into consideration what I said above about the pouch conforming to the button, it would seem almost unnecessary to be concerned about pouch-to-button clearance. Here we get into physics, which I've never studied, but I do understand some basic concepts. In this case, it has to do with the square area of free space inside the pouch well and the dynamics of the pulse of air that will inflate the pouch. Since the pulse that inflates the pouch starts out small and increases to its maximum as the hole in the trackerbar is opened to 100%, and remembering that the pouch has its maximum power from time zero, and further taking into account that the pulse has to first overcome the air leakage in the well -created by the bleed- before the pouch starts moving, it would be most advantageous to have the valve button as far down in the pouch well as reasonably possible.