***Note from Migration:***
Original Title was: Instead of providing the Python dict fme.featuresWritten indexed by feature type, give us a nested dict indexed by writer name and then feature type
... this way we could extract the number of features written even if two writers have feature types with the same name.
