It relies critically on estimating the number of asymptomatic cases. To do this accurately you need either the PCR test to report a positive for a long time af

It relies critically on estimating the number of asymptomatic cases. To do this accurately you need either the PCR test to report a positive for a long time after contraction, or you need a serology-based test. They are basing the asymptomatic case % from the Diamond Princess data and assuming that all people who tested positive where the only ones who had cornavirus. Is this warranted? Has anyone seen any data on the period of time after contraction during which the standard PCR test will return positive?