[jsnell]

1/1000 was not realistic when the preprint was published, let alone now. It showed that one could fit a bunch of curves onto 14 days of data on deaths. Which is true, but also utterly unintersting. The reason that sorry excuse of a model got so much attention is that the authors did not stop there.

They also made the totally unsupported assumption that one of the lower end curves matched reality. (That's right. They actually made exactly the kind of assumptions you're accusing others of, rather than just argue for antibody testing).

The problem is that when the study was made, there was a lot more data available than just that 14 days of deaths from two countries. And a ton of it was totally incompatible with their modeling.

Here's some of the conflicting data points as of ten days ago: https://news.ycombinator.com/item?id=22698584