The PCR test is not a bad test for detecting these things. Quantitative PCR (qPCR) is useful for relative measures, and a non qPCR can be useful for just detection. The problem with the coronavirus PCR test as a test for Covid-19 is twofold.

1. It is prone to errors (false positives and false negatives) even when done properly. Regular PCR has an error rate in the 5-10% range (so 90-95% accurate) for lab samples. I am not sure what the error rate is from samples in the wild, but its likely higher.
2. When cycled above 25-30 the false positive rate increases dramatically with each cycle. The protocol from the WHO set the recommended cycle rate at 45, for a non qPCR test. Such a thing is so far beyond ridiculous its difficult to express how fraudulent it was obviously intended to be.

Neither of these precludes it from being VERY USEFUL in a lab setting for detecting the virus (especially a qPCR test). Conflating these two tests: qPCR in a lab setting, done multiple times and an overcycled PCR test in the wild is mixing apples and oranges and expecting a banana.