And it’s not just an abstract “economic” cost on the X-axis, ultimately it also comes back down to quality of life and even a cost in lives. Actual people losing their businesses, losing their houses, etc.
I can’t help but think of Asimov and Three Laws. Taken to its extreme, robots interpret the law of “not allowing humans to come to harm” in a way that they ultimately enslave humanity.
First, I think the nCFR is fundamentally and significantly overstating the risk to the general population because orders or magnitude more people already have it than have been tested positive. Second, I think the treatment for COVID is pretty basic. With a concerted effort to deploy temporary facilities across the US we could add sufficient ventilator beds for a quick peak in cases and save trillions of dollars in economic damages.
However, this guy (who, by the way, is just a developer who has given "a few days of thought" to this issue) does not seem to know anything about this issue but still proposes a conspiranoic theory that HN entertains way too much.
I took his point to be simply that flattening the curve does itself have significant associated costs. That a maximally flattened curve is not necessarily the best approach.
For people who rely on public transport to get to work and who can't work remotely, I think they would argue that the cost / day of flattening efforts itself also begins to rise significantly. That is, the first week or two are less impactful than subsequent consecutive weeks. And that cost is borne widely. That increasing cost may or may not balance out the failure to have the cost axis take into account the height of the curve.