(1) As an additional systematic check, could you compare the results obtained from a "counting experiment analysis" with those you obtain here from a "shape analysis"? That is, normalize the B and Bbar samples individually, count the number of signal events and use them to measure the direct CP violation.
->
(1) As an additional check, could you please compare the results obtained from a) a fit which normalizes B and Bbar separately (a "shape-only" measurement) b) a time-integrated measurement, i.e. counting the # of (signal) B and Bbar decays with the nominal result (which implicitly combines the above two results), and verify that the results obtained are consistent.