I think the mistake that many people keep making is trying to nail down a single IFR. There isn’t one. The fatality rate depends heavily on the demographics of

I think the mistake that many people keep making is trying to nail down a single IFR. There isn’t one. The fatality rate depends heavily on the demographics of the infected population. IFR of a entirely young demographic would be below 0.01%. IFR of a nursing home would be 3 orders of magnitude higher. If someone shows me a low IFR I can safely bet that they surely aren’t finding 50% of their cases in their nursing homes. Conversely, if 50%+ of your cases are in nursing homes, your population IFR is going to look a lot worse than if you had kept that relatively small population safe.