There's no rubrik for grading scientific correctness. Just because controlling variables is more difficult doesn't mean we should abandon the scientific method and just talk out of our asses.