The test only checks for active virus. FOr a real evaluation, we need a test for anti-bodies.
Yes I realize that. If a test for antibodies exists, then we should be using it to understand where we are in the course of this epidemic. But until that test is available a large random test for the active virus will allow us to estimate the number of current infections within the total population and from there get a better understanding of the true death rate. The 2-4% death rate numbers thrown around are simply the number of deaths divided by the number of positive tests. But the current number of positive tests are way under representing the actual number of current infections and thus results in a very large over estimation the actual death rate.