Fixes to DC geometry
Created by: raffaelladevita
Accounts for:
- the mini staggering of the guard wire used as a reference in R3;
- the shifts to the wires due to the finite size of the feedthroughs that hold the wires in place.
For the latter only the shift to the wire midpoint is used while the change in direction is for now ignored because somehow tracking vertex resolution become worse when that is used. The current hypothsis is that is just a mathematical issue appearing because reconstruction is still assuming the stereo angle is 6 deg. The effect should be anyway negligible because the max change to the stereo angle is +/- 0.013 deg, corresponding to shifts of the wire endpoints of 10-30 um.