These "statistics" are not credible. More than 0.1% of all Americans (340k) have already died of COVID-19. If the IFR were 0.15-0.2%, more than half of the population would have been infected already and the spread would be pretty much over. Instead, seroprevalence shows that about 20% of Americans have been infected, meaning we have 600k or so deaths to go in the absence of controls.

As for "99.96% of nearly everyone who gets the virus lives" this is weaseling even given their low IFR estimates. It deliberately excludes those most likely to die! This doesn't look like good faith. An article about the harms of isolation doesn't have to, and shouldn't, minimize the facts in this way.