Draft: Improve rd multilepton lines

Starting with a list of things to improve:

• Add intermediates to B->4mu line same as for B2EEMuMu, generic four lepton builder based on two dilepton builders
• Add intermediates for B->4e line more difficult, it creates electron candidates very simple while dielectron objects (which are used in other lines e.g. B2EEMuMu) are way more complicated, is this intentional and the difference understood? Can we change also B24e to align with the other one?
  • This makes tupling simpler and control plots/cross checks
• Change BPV functors in builder with OWNPV
• Remove isolation variables, tagging_particles for persist_reco=True lines as these are not needed
• Think about TBCs in rdbuilder_thor.py (more of a general thing), e.g. trchi2dof_max, trghostprob_max
• Rename make_b_4l keyword MAXDOCA with MAXSDOCA
• Add sprucing (?), see comment here
• Consistent naming of keyword arguments, e.g. here and here
• Combine make_b_4l, make_b_2mu2e and make_generic_4lepton_LLP into one builder function
• Reconsider Kaon PID cuts
• Consider renaming Nmu_builder.py to b_to_multilepton_builder.py to be consistent with other builder files and because it is not only "n times mu" what is build here
• Consider to rename lines e.g. LL like here as there are no not long muon lines it could be confused with
• Why min_PID_Mu=1 at one point and min_PID_Mu=0 on another
• Why min_PID_E=3 at one point and min_PID_E=2 on another
• MAXSDOCA cut has unit in most other use cases throughout Moore, is this intentional that here it has no explicit unit?
• Add downstream LLP lines with various topologies LLD, LDD, DDD