The curve flattens naturally at some point, always after “something” has been done. That does not mean these two things are related to any significant degree.
The number of of actively infected non-quarantined people is relevant. Take a look at RKI-dashboard for Jena. This Jena paper is a serious fake.
I think mask measures are not working correctly - all the masks are way too dirty - effects are most likely going in the wrong direction.
Getting good data to evaluate measures currently isn't easy. Some curves using "reporting date" look seriously different than curves using "date of death".