I hadn't looked at the US fatality rate for awhile, as the numbers earlier were so skewed by the early hot spots, like New York, when the lack of testing made the number somewhat meaningless. A lot of people died without getting a test.
The number still depends on the rate of testing, so it is not a perfect measure of what % of people who contract the disease actually die, but it's probably closer to the "real" number than it was back in the first few months, and... 3.1% is pretty high!
