And of course Covid tests are primarily mucous membrane based which is going to be inherently harder to evenly test compared to a blood sample where viral load is pretty evenly distributed.
At least for blood based diseases, detecting viral load via PCR testing is to such a sensitivity that if there's essentially any active viruses out and about or any active viral RNA floating around in cells then the tests come back positive.
And with sustained antiviral use testing is less about "do I have it/will I become contagious in the near future" (like coronavirus) and rather "is my antiviral regimen still killing the virus faster than it can wake up from latent genetic material sleeping in DNA".
The former is a timing problem from one shot testing the latter is monitoring a steady state to track that the treatment remains effective and when it ceases being effective there's a lag time between viral load being detectable and sufficient viral load to be meaningfully contagious.
By contrast, testing with rapid tests is mostly about seeing why you're sick before you go somewhere, and those tests probably don't need to be as sensitive. The hospital has capacity to treat people because of the PCR tests and so we can manage if some of the PCR tests are false-negative. Most people are vaccinated so we can tolerate some Covid-positive people mingling in public places now.
Tests aren't magical, and there are different tests with different capabilities. I'd suggest you learn about sensitivity versus specificity and learn how to compute the false positive, false negative, true positive, and true negative rates. It might be eye opening for you.