Don't think it's the number (2, or > 1) that matters, it's that there are things not yet tested, that it successfully predicts (and that other models/theories fail to predict).