Despite the media coverage, this study is completely theoretical. It is a computer model that shows that multiple sets of assumptions on the infection rate and the hospitalization rates would explain the current number of deaths. For example, if the severe illness rate is low, it implies many more people must have had it. In the study it also shows that it could be the case that only a fraction of the population has had it. That didn't make the headlines. The real point of the paper is that without serologic testing we won't know. But do you really think 50 percent of the UK had this? Maybe more had it than we think, but 50 percent is not very consistent with the testing we're seeing.