Do they account for potential variations in gravity where they've been doing the measurements? Given how ridiculously sensitive the equipment is the position of the moon probably has an impact on the readings they're getting.
Down to eight significant figures it is an issue (effect of the moon's gravity on the surface of the earth peaks at 1.1 × 10−7 g), but you can always point the normal of the axial plane of the balance in the earth relative right ascension direction of the moon to avoid this effect (regardless of the moon's relative declension, δ).
