You're confusing case fatality rate (detected infection denominator) with infection fatality rate (total population denominator). The study I cited was (broadl

You're confusing case fatality rate (detected infection denominator) with infection fatality rate (total population denominator). The study I cited was (broadly) CFR. That Lancet study is (broadly) IFR. If you proactively surveillance test 13.5% of Madurai's population, you're going to get a different class of people in your denominator, and your calculated mortality is going to decrease. The BMJ study is probably the most broadly compatible, which leaves us at 2.5% (non-Delta) or 4.1% (Delta), which appears to be case fatality rate? If we want to argue apples:apples, we should use EMR-based studies that attempt to calculate IFR, as at least they're reasoning with the same population composition. This [0] has SARS-CoV-2 at somewhere between 0.26% and 0.83% IFR. Which for context, is compared to respiratory flu at 0.0088% (estimated total population EMR due to flu) [1]. So we'd expect about 29.5x as many people to die from SARS-CoV-2 (using the high end estimate of flu and the low end estimate of SARS-CoV-2, to generate the lowest possible multiple). If you want to, you can probably dig European numbers out of this excess mortality calculation [2], although that's all-cause mortality, with no attempt to isolate cause. [0] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7543961/ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7543961/ [1] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5935243/ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5935243/ [2] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7346364/ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7346364/