Multiple-comparison correction across many evals

[delcenjo]

Running the gate over many slices or many evals inflates the false-positive rate. An optional Benjamini-Hochberg or Bonferroni adjustment over a set of results would keep the overall error rate controlled.