While we can obviously do a good job on building a precise detector and aligning planes to below 10 microns, this study raises concern that some of the GEANT parameters may cause systematic biases of over a hundred microns.
I hope this would be the first of a series of investigations of the various parameter settings in GEANT 3. The sizable and systematic nature of the effects of this parameter raises concern that other parameters may result in similar biases. In the future we might want to further investigate the effects of some of these settings on the Michel parameters to quantify their effects.