It's not possible because the probability of incorporation of B and C into the film from the precursor gases is different. This depends on the probability of B or C attaching and sticking to the surface of the growing film, which depends on temperature, and perhaps also pressure. CVD kinetics tend to be quite complex so you'll probably need to measure the B concentration directly on at least 1 sample. Linearity of B content with trimethylboron is not guaranteed but probably an ok assumption.