This is an important question without a singular answer.
Presumptive positives can be the result of:
1. low starting viral quantity
2. presence of RNA fragments that happen to match one (orf1ab) but not both targets
3. low levels of contamination
4. non-specific amplification during later cycles (stochastic effects occur; this is why the test has a cut-off at Cq 40 for positive results)
...or a variety of other factors. For most assays, there isn't really a presumptive positive category. However, because of major risks to the individual and the public due to a false negative result, the FDA and CDC want to be very aggressive in treating anything as a positive that has even a slight chance of being so.
In our practice, we always rerun these samples after re-extracting the raw sample, and the vast majority of presumptive positives come back as negative on the rerun. The presumptive positive allows us to err on the side of caution, rather than to risk potential false negatives.