If there is that much variation in manufacture, I can see where the two arms (relative to their respective semiconductor arcs) perform some type of averaging, or some other mathematical function, such that the relative distance between the two arms along their respective arcs are the calibration points.
I don't know if that came across intact, but I am imagining the distance between the two arms themselves are the calibration...i.e bending them closer together or further apart might be the method of calibration.
The problem I have with this is that a professional body man can spray .875 mil of paint by feel/sight. Surely the machines Bosch uses to produce this part can lay down a strata +/- 10-20 microns. I cannot conceive that one of the founding ISO 9000 companies can produce a part of such varying consistancy that this can be an issue.
Of course, all this is academic at this point as the only source for this part I have found so far is this website
and after checking tonight, the only person at work who speaks Russian left in January.
I would love to test this theory if anyone has any sources for the Bosch part 3-437-224-035 (the number on the original is 3 437 224 015), but it looks like I will be ordering a new air meter this week.