Should we narrow the focus to cases with same binning (simulation can always be set to real detector binning), or consider how to compare data with different binnings?
Is there a way to compare event data that was never in a histogram to begin with?
Assuming we find a systematic way to do this comparison, how do we implement it? As a little library we all use in ViNYL?
As a first step, perform a literature survey on established best practises for sim-exp benchmarking and comparisons.