Several problems with this analysis:
- Using confirmed cases as the denominator significantly overstates the fatality rate, because a large share of the total infections is never confirmed
- "All age groups over 18" elides huge disparities between the younger and older age groups. It also obscures large differences in risk between those with comorbidities vs those without. The CDC does not make this easy to parse out, but in Canada after their first wave, "100% of the COVID-involved deaths under the age of 45 as of July 31 had at least one other disease or condition certified on the medical certificate of death." https://web.archive.org/web/20210219195135if_/https://www150...
- New York as a region had one of the highest fatality rates of the entire pandemic
- Covid is a "one-and-done" risk as far as we know, while the risk of a car accident is repeated year after year for as long as one drives on the roads. Even in the absence of vaccines, the risk of death from Covid is unlikely to remain as high in 2024 as it was in 2020.
The most sound way to run this sort of comparison is something like "how many in risk group X (say, people aged 30-39 with no comorbidities) died in 2020 from Covid vs how many in that group die in a typical year from auto accidents?"
If you want to say that overweight 60 year old teachers with diabetes should not have been forced back to in-person instruction you'll get no argument from me.