It's precisely because almost all gravitational dynamics are controlled by G*M (for some M) that G is so poorly-known. We're forced to do terrestrial experiment

It's precisely because almost all gravitational dynamics are controlled by G*M (for some M) that G is so poorly-known. We're forced to do terrestrial experiments, and they're extremely difficult. G is very seductive. It seems obvious that we should be able to measure a fundamental constant of nature to better than 14 ppm (or, if you take the scatter of current measurements, 100s of ppm), but it's really really hard.