It should be noted that gearbox rebuilds can be made a lot easier by dispensing with the need for and expense of a depth gauge and truing plate. The shaft shimming can be set empirically with good accuracy via trial and error. In using a truing plate and depth gauge, I got variations of .02-.04 mm depending on where I rested the gauge and how hard I pressed on it to hold it steady on the small bearing race surface available. Getting the gauge foot to set true on the inner race was a nosebleed.
After reading every forum entry and tech article available on the topic on the web, it became evident that the _minimum_ shim clearance was not crucial as long as the shafts spun freely with zero preload, but just zero preload, no slop beyond that. Being that my margin for error using the depth gauge was almost that amount, it was possible to exceed recommended clearances given variations in gasket crush, running temperature, shim size steps, etc. The gearbox case is always going to expand at a greater rate than the steel shafts, so slop will always be introduced. If my measurements were towards the loose margin on assembly, the shafts could end up with more than .05mm at operating temperatures.
So I installed the shafts, one at a time, using thinner and thinner shim combos, until the shafts spun with a light amount of drag and then went .02mm thinner on the shims. The layshaft was turnable with a screwdriver via the output shaft hole minus its seal. This procedure involved removing the case cover many times (blow dryer for heating the cases) and care was taken to iinsure that it was tightened down equally each time using the clutch settings on a battery drill. As the shaft became movable with thinner shims, it was easy to gauge the lessening amount of drag when rotating the shafts, each in its turn, and when they turned totally free, it was empirically/tactilely evident.
The gearbox has several thousand miles on it and is quieter and smoother than ever before.
