Preprocess: Remove reliance on Peak-to-Peak cuts

[msilvafe]

Currently we are using a peak-to-peak cut that looks at the kurtosis of the distribution of the peak-to-peak of each TOD and cuts outlier detectors until the kurtosis is below some threshold. I see this as a bandaid that is mostly cutting detectors with jumps that aren't being properly flagged + fixed. Once we are more confident in our jump finding + fixing being tracked in issue #975 we need to re-evaluate the need for the peak-to-peak cuts and test if we can remove this step.