No assumptions about pion tagging so far, and no neutron tagging. In the future will look into this, assuming various efficiencies and cutoffs.
Rik says at some point pion mass energy is visible. In liquid argon especially might see this when pion decays. There is the potential refinement to include this in the future, but not in there now. But in gas can hopefully count pions.
Removal energy=amount of energy it takes to get the struck nucleon out of the nucleus.
Asher points out these biases are also very different between the various generators. Can use the to motivate a fake dataset.
Stephen points out that some of these are models are known to be wrong. NuWro is the least wrong. If you add DUNE baseline Genie model, it looks more sensible. Also make form factors different for different resonances.
Tom says generators might have different ways of labeling and accounting for resonances when there is an absorption.