Some unit tests are using loops generating random values. In some cases, this could be replaced with proper fuzzing. There are even cases where the fuzzing tests have already been written.
I propose we remove the random and/or looped unit tests when an equivalent fuzzing tests exists.