I understand there is no baseline, but consider the example data I posted, at 20%, the sample failed, which means it was > 50ng, and at 16% it passed, meaning it was < 50ng. Therefore, the concentration in the original sample is between 50/20% and 50/16%, or >250ng and <312ng. If you then tested 19%, 18%, 17%, and found it failed on all of them for example, then you know the original sample is >50/17% = 294ng, So your new range would be 294 to 312. If you wish for a more accurate result, then repeat the experiment.

Everything has some error, the thing to figure out is how much. Say a test strip can measure 50ng +/- 10ng, then the calculations are still sound, but need to incorporate this error. This error can be determined by repeated testing of a known concentration, this would require a GC/MS. But for the person trying to figure out if they're at 1000 or 100, this is unnecessary.