My first thought was that we only know Cavendish's constant to a little over 4 significant figures, so how could this be right? The relativistic effects at Earth's surface would change this by only ~10^-8, so I think the challenge in refining the Cavendish gravitic constant lie elsewhere.