> Two things: first, you don't need every hospital to be overwhelmed for it to constitute an issue.

Of course, but you need a significant amount of hospitals to be overwhelmed to cause significant excess death. Remember, we're trying to minimize excess death of all causes, not just COVID-19.

> Second, every successful preemptive measure is by definition an overreaction. Perhaps you don't see more overwhelmed hospitals precisely because lockdowns are effectively preventing that.

Sure, but that insight doesn't really help. Maybe that's true, maybe it isn't. Given that neither Brazil nor Sweden had a lockdown and given that neither of their healthcare systems collapsed as some models predicted, my guess would be that a lockdown isn't necessary to prevent such a collapse.