In fairness, it's good to be willing to try a different approach from first principles. It's very easy to get locked in by your preconceptions about what works, and as a result miss out on better approaches. The thing is, you also have to have the humility to admit that the new approach is not turning out to be better, and uh... Elon doesn't have that to say the least.
I agree to an extent. If the new approach shows some kind of merit, and that merit is not going to destroy one of your primary requirements, go for it. But doing it for no reason doesnt make a lot of sense.
