Yes, although usually when people are making these arguments they're imagining optimization across different entities and through time (relative to foreseeable events / probabilities). It then becomes much easier to write off conflicting higher-level subjective goals as not being actually valued in some imagined universal sense. The fallacy is rejecting the stickiness of human-made structure, even though that structure is all around us.

It's hard to argue that as a society we should construct systems that have any goal that isn't optimising (in this example) services. If that aligns with maximising profits for shareholders, then great, if not then the system is not working as it needs to. IMO it's never really about free markets per se, but about competition. If the market is not delivering effective competition, then it should be made to do so. Competition is what aligns shareholder interest with customer interest.