[QEE]: loosen pre-vertex mass cut on Zs following !4151

Aim and changes

Z cuts:
master:

• pre-vertex mass cut of min_mass (nominally 40 GeV)
• no post-vertex cut
• This causes low-mass (post-vertex fit) dimuon candidates to propagate through the dataflow before being removed by analysis level mass cuts.

!4151 (merged) i.e. qee_upgrade currently:

• pre-vertex mass cut of min_mass (nominally 40 GeV)
• post-vertex mass cut of min_mass (nominally 40 GeV)
• Removes the unwanted low-mass (post-vertex fit) dimuon candidates

However, due to a non negligible pre-vertex mass resolution, cutting at 40 GeV pre-vertex fit may remove real events with post-vertex fit > 40 GeV that we wish to consider. To prevent this harming signal we make a looser pre-vertex mass cut followed by the actual post-vertex mass cut we want for the Z dilepton.

Thus this MR, Z cut is:

• pre-vertex mass cut of min_combination_mass (nominally 30 GeV)
• post-vertex mass cut of min_composite_mass (nominally 40 GeV)
• Keeps all the signal we want while maintaining decent throughput from combinatorics with the pre-vertex mass cut.

Effect on BW

Effectively no change, estimated to be permille level or less rate change on the relevant lines. (Considering !4151 (merged) was an immeasurable reduction, this is an even smaller change so can safely be ignored)

cc @rjhunter @mvesteri