I would take that number with a grain of salt too, as there is a lot of asymmetry with how and when that data is reported. There may be a large number of "recov

I would take that number with a grain of salt too, as there is a lot of asymmetry with how and when that data is reported. There may be a large number of "recovering" cases that are not completely closed yet but have essentially zero chance of death in the future. The definition of recovered seems open to interpretation while death is not, and so the deaths would have a more immediate reporting period, as well as a shorter average interval of time since infection started.