For it to be "obviously false" you seem to be assuming that the comparison in energy use should be between the Bitcoin network (miners and all) and bank infrast

For it to be "obviously false" you seem to be assuming that the comparison in energy use should be between the Bitcoin network (miners and all) and bank infrastructure ("a few database servers per bank" plus "people operating phones ... and other things"), whereas I think that the OP (and others) believe the comparison should be between Bitcoin-the-currency (still miners and all) and USD-the-currency (many dependencies!). That's much harder to cost and argue which is why I'd like to see a serious analysis on this I realise that using the phrase "the banking system", as the quote does, makes it ambiguous.