Earlier quoted context omitted.
CO2 emissions are Θ(P). I don't see how you could say this "does not correlate directly to CO2 emissions".
Its not complicated. It is not necessary for crypto PoW to be powered by CO2 emitting fuels. It can be powered with renewables. Therefore the power usage of Bitcoin can't be used to infer its environmental impact. You need to know how much of that usage was on the back of fossil fuels.
If base load were to drop due to crypto miners being turned off, the energy with highest marginal cost would be turned off, which means fossil fuels, not renewables.