There are a few odd ones in the list:
4 {}
3 ...
3
Are those real or artifacts of how you collected the data?
It’s interesting that so many highly-used features do not have information about them in feature_gate.rs. What is the difference between the ones in that file and the ones from the tidy script?