To say something about my background, I'm trained as a psychologist, and I work in the field of psychopharmacology at my university.
In terms of oral fluids, I don't think that there is a single relationship between THC concentration and oral fluid relative to blood. It depends really on the device that people are using. There are many differences between the devices themselves that can lead to a different level of THC that has been traced in all fluid. Also the oral fluid devices are usually not able to detect THC levels for a very long time, so there's really a narrow time window of about three to four hours that would enable you to pick up recent THC use. Even more important, I think, is the real cut-off value that these oral fluid devices are operating by. They can vary quite a lot between devices.
If the sensitivity of such a cut-off level is really high, meaning that if the cut-off level is really low, then the rate of false negatives would be very low as well. If the cut-off level were higher, for example 30 nanograms or so, then I predict that the false negatives raised with oral fluid devices would easily be up to 50%, even if people smoked a cannabis cigarette 30 minutes before.