Track pulls are not-optimal

[Zehvogel]

A part of this was already fixed in AIDASoft/DD4hep#1414 however, the Omega pull is still too big (sigma 1.7 instead of 1.0).

We would need:

• A more detailed investigation of the Omega pull distribution for multiple regions in phase space (potentially using https://github.com/key4hep/k4DetPerformance)
• Someone to actually figure out where this goes wrong ;)

Here is the current status for 10k 1 GeV muons at theta = 89 deg copied from the PR above