Characterize peak suppression

[rmflight]

For a peak across scans what is the fraction of scans with:

• missing or zero value
• zero values instead

This provides measures of "peak suppression" and error introduced. Histograms of these fractions should give us an idea of how many peaks are seeing missingness and how many have error introduced and how much.

This came up in a discussion with @hunter-moseley about examining error from other methods.

[rmflight]

This is actually handled by the differences of NAP - intensity and what happens with XCalibur especially.