Right now Bitcoin is an additive to the financial system, and actually promotes more usage of fiat. So if you were to compare the manifold energy consumption to Bitcoin, you need to somewhere also calculate those other effects.
I can't find a reference, but there's a good phrase that talked about how for every new innovation that comes out and is going to "completely replace X", the actual equilibrium always turns into a hybrid of the legacy systems and the new systems. Much like we've seen with Bitcoin.
So at least at this point, I find your argument disingenuous, because it doesn't take into effect the interplay of the systems.