Earlier quoted context omitted.
When the alternative is simply just using Proof of Stake and lowering the energy consumption by an estimated 99%, yes, we can be judgmental of how others use energy in this specific case. It's not Bitcoin/PoW or no cryptocurrencies at all, that's a false dichotomy, if that's what you're implying.
PoS lacks fairness. To participate in PoS, you need an existing stash of coins, which is inaccessible for many individuals from countries where cryptocurrencies are outlawed. Not forgetting to mention that participation in consensus would indirectly require going through KYC. PoW, on the other hand, is akin to buying Bitcoin with electricity - a borderless natural resource.
Those three things would give a government interested in banning cryptocurrencies plenty of leverage to disrupt mining operations in their country.
All they would need to do is regulate the import of ASIC mining hardware; sinkhole any traffic that looks "bitcoin-y" (damn the collateral damage, we're fighting fascist communist pedophile terrorist drug traffickers here!) and similarly block access to any known mining pools.
They could go even further and introduce severe and draconian penalties to anyone producing, possessing, or using cryptocurrencies (which of course would be selectively enforced).
At this point, once you factor in the real-world element, the fact that, in theory, you only need electricity to produce bitcoin becomes such a small part of that equation.
I think at that point why not use PoS (or better yet, a DPoS scheme), if there isn't any real-world benefit to PoW? I think the tradeoff of introducing a small amount of revocable trust in exchange for RADICALLY reduced energy consumption is pretty clearly worth it, given the real world constraints.