Gen muons from taus
It is customary in DY analyses to reject muons (and electrons) form the decay of tau leptons. A flag for this needs to be computed in the Ntupliser and added to the trees (gen muons are still needed for the Rochester correction).
I'm not sure how this additional flag would fit in the current class hierarchy. The perfect solution would be:
BaseMuon(p4, charge)
GenMuon(originatesFromTau)
RecMuon(selectors)
This also looks a bit complicated (we're slowly reinventing the CMSSW class hierarchy).