Numbers depend on the age. Bulk of deaths are concentrated in older populations. CDC recently updated estimated infection fatality rates for COVID. Here are the updated survival rates by age group:
The CDC updated their pandemic planning scenarios[0] based on earlier studies in Europe plus some educated guesswork using data collected from the US through August 8. The numbers which you represent as the CDC’s sole official estimate of true IFR are only one of three possible sets of parameters that they provide for scenario modelling. Furthermore, the footnote on these estimates states “The estimates for persons ≥70 years old presented here do not include persons ≥80 years old”, since the underlying study they base the model on assumed that CFR = IFR for people ≥80 years old. Why they label this 70+ instead of 70–79 is beyond me, and I have no idea how sound the methodology is that they used to derive these estimates.
I would also mention that there are studies in peer review on patient populations in the United States which suggest IFRs closer to CDC planning scenario 4/5 than 3—for example, this one from Connecticut[1]:
We all hope for the best—that the lower bounds are true—but I think we should also be prepared for the reality that the upper bounds might be the correct ones, and act conservatively.
Also, you know, all IFR estimates assume that patients will actually be able to access care. Without hospitals, IFR approaches IHR, and the hospitalisation rates from the Connecticut study are grim: 0.8%, 2.68%, 3.09%, 12.43%, and 79.89%.
0-19: 99.997% 20-49: 99.98% 50-69: 99.5% 70+: 94.6%
Edit - typo.