Accuracy is relative I guess but my understanding is that the cycle counts used most commonly thus far for COVID-19 (30-40 with a possible lowering to 28), there would not be very many false positives. A test in the 35-40 count range would be more likely to pick up early infections and especially recent recoveries than a lower cycle count though.
I read an article somewhere that mentioned how it was desirable to use a higher cycle count among those hospitalized with symptoms as a positive result would not likely be because of a false positive, a recent recovery or early infection in that case. False negatives are actually far more common that false positives and you don't want to get a false negative with someone who has severe symptoms. On the other hand, you want to try to avoid detecting recent recoveries as much as possible (which apparently is more commonly detected than early infections) in the wild so you would want to use a cycle count on the lower end.
I think there is plenty of reason to fear viruses. At least certain ones. I wouldn't want to try my luck with Ebola for instance. Of course, COVID-19 is no Ebola. But there are plenty of nasty, dangerous viruses that can do bad things to you. Fortunately, most of them don't spread too easily or are otherwise not too difficult to avoid.
As for the numbers an graph for Cambodia, most likely somebody's honest mistake somewhere. At least I can't imagine why somebody would want to fudge the numbers for Cambodia. But as Mark Twain said, "There are three kinds of lies: Lies, Damned Lies, and Statistics". Statistics are great but you have to know the details about what was measured and how the data was collected to draw real conclusions.