PCR is technically being used to diagnose active disease, but as been well elucidated elsewhere is not a great test because of the over amplification and while it may demonstrate presence of the virus in a sample, it may not quantify if sufficient to cause an infection. Its the old “true - true but unrelated” foil.
There is a narrow window of actual detectable antibodies usually around 90 days. After this — it gets to T-cell immunity with memory. So It is correct it is not a good indicator of infection unless caught within 90 days of symptoms.
Thanks. I only had the test last month, because I was concerned about some unusual symptoms I had noticed many months earlier. But they were probably just due to disruption of routine, lack of usual exercise, etc.