one directly contradicts the Stanford and LA ones. The details of the NY study haven't been published or analyzed, but that one seems reasonable from the numbers shown (specifically the % values on the data for upstate NY puts a bound on a false positive rate). Stanford was estimating that the undercount was in the neighborhood of 50-85x, while the NYC data put it at something like 12x. That's a pretty big difference, and even if NYC had 100% infection their fatality rate would still be higher than what Stanford was estimating. The Stanford data is highly suspect. Their false positive rate (that they acknowledge) would mean that about 33% of their positives are false.

The common theme is that the infection rate is undercounted, and I don't think anyone at all disagrees with that. But whether it's a factor of 10x, 50x, 85x, or more is significant).

I haven't looked at the other studies (or read peer critiques) close enough to really have many thoughts.