Something buggy in 0.preprocess-sites/0.prefilter-features.py

[ErinWeisbart]

Haven't tried to track down the details, but both I and Meraj have encountered this week with totally different datasets a situation where 0.preprocess-sites/0.prefilter-features.py doesn't contain the correct set of features.
I believe mine was a situation where I had run it once, then change some things in the config and triggered it again and despite 0.prefilter being set to true, the feature_prefilter.tsv did not update appropriately.
Potentially Meraj's was a situation where there were different plates with different features and he set the config to only use one of the plates but it generated from the set of features from the initial plates, but not totally sure this is what happened.