The opposite actually. When the dag becomes increasingly complicated, task_dict will start to clutter, but in a much more manageable way since it's in dictionary form. DAG.py files are kept clean for front-facing purposes as well. For our team, this allows data analysts who are not so technical to easily deploy DAGs