True, but no estimate is lower than 0.3%, which is still very high (600,000+ dead in the USA). The next step in walking back the fatality rate of course will be to define an IFR that's "net of people who were probably going to die soon anyways".