As I understand it, the quick antibody tests are the unreliable ones. For these tests, the blood will be collected at home and sent to labs where i would suppose reliable tests will be done.
Also, the tests will begin next week and the results will be published end of April or beginning of May.
In principle even if the tests are unreliable in individual cases, if the false positive and negative rates are known sufficiently well it ought to be possible to use mass results to get an accurate estimate of the prevalence of what they're supposed to be measuring.
He said they tested people for immunity (my interpretation was antibody tests, maybe i was not correct in the actual type of test - but they are testing for "immunity" (antibodies?), somehow).