The formula is correct.
The main feature of antibodies is that they have two binding sites.
Assuming that you use the antibody as ligand, the two binding arms are treated as independent binding sites on the sensor surface. Therefore, each antibody contributes two binding sites. Dividing the Mw of antibody by 2 is the same as the formula below.
And, when using an antibody as analyte the second binding site will not generate any response but shifts the kinetics to a higher overall affinity. In addition, it is impossible to say that all antibodies bind with one arm or all antibodies bind with two arms because this depends greatly on the ligand density. In conclusion: you cannot calculate the functional activity of an analyte when it is a bivalent molecule.
Kind regards
Arnoud