I think you are right but you are not meant to detect them in the open enivironment where workers are exposed.
I understand that benzene was no longer brought into these sites so that any benzene detected must have been produced in-plant by chemical reaction, but again should not be detectable.
The cancer-inducing nature of materials is also a function of the amount of material present.
I remember in another life when i had to read a learned paper about a food coloring called "Herring brown B." The paper supposedly showed "Herring brown B" to be a carcinogen.
However, the amounts of "Herring brown B" fed/injected into rats were enormous, at some significant fraction of their body weight over some period of time (I am sorry but the details are forgotten).
I can't even remember how Herring brown A and Herring brown C compared.
But I remember thinking at the time that at that amount of exposure almost anything could show as a carcinogen.
The chemicals used in fabs are dangerous, especially arsenic. Hydroflouric acid is also very dangerous. The key is to use them without exposing the workers. I guess this study was measuring the concentrations in the air, so the workers are exposed. I wonder how the US fabs measure up.