Allowing beamspot smearing and disabling beamspot constraints in PV fit

[saracreates]

BEGINRELEASENOTES

• Set beamspot (BS) smearing to true, justification shown in this issue
• Set beamspot constraint for PV fit false as default (WARNING, this is NOT a permanent solution. With BS constraint the PV are too constraint. Someone needs to investigate why. But if we do not constraint the PV fit with the BS, the resolution is as expected. Only the resolution of $y$ is influcend by this change as only in $y$ tracking dominates the resolution, otherwise it's the BS anyway. For thorough explanation, see this issue

ENDRELEASENOTES

[saracreates]

This PR is not the solution to the problem.

• DO NOT set the beam spot constraint to false. We need to apply it to get reasonable results. See Transverse impact parameter discrepancy between full and fast simulation #61
• I would still suggest keeping the beamspot smearing to true.