Sherpa seems to duplicate taus for born leptons, i.e. status == 11. This should remove this. Checked for mc21_13p6TeV.700622.Sh_2212_Ztautau_maxHTpTV2_CFilterBVeto and mc21_13p6TeV.700600.Sh_2212_llll. Is a bit tricky in that the second set of four leptons for ZZ sample can be tautau, ee with tau having status == 11 and electrons status == 1. Jira ticket ATLASG-2516. @egramsta and @mhank may be interested.