The color of gold is better explained by an ad-hoc combination of SR, GR, and QM, than by any one of them alone. SR and GR explain the planets' motion better than Newton and Kepler alone. QM explains pi bonds in molecules like benzene better than classical chemistry. GR compensation is required by GPS satellites and receivers.
So it doesn't have to be a combination of GR and QM, but it does have to be something which is at least as good as GR and QM respectively in their existing domains of explanation.
So it doesn't have to be a combination of GR and QM, but it does have to be something which is at least as good as GR and QM respectively in their existing domains of explanation.