I don't think its premature when we have stuff like the model of Oppenheim et. al. which appears to predict stuff which we can observe from astronomical data.

I don't think its premature when we have stuff like the model of Oppenheim et. al. which appears to predict stuff which we can observe from astronomical data. This paper [1] also discusses some experimental tests of their model which are not too technologically insane (you don't have to put a planet in superposition). Pars of the parameter space of their model are already ruled out by stuff we already know and more can be excluded with a combination of deriving better theoretical bounds (their analysis is quite coarse to simplify things) and better experiments. I don't think its very meaningful to talk about any "measurable discrepancy between GR and QM" in any case - we simply don't have any predictions in many cases where both QM and GM are very relevant. General relativity doesn't tell you want happens when you put a particle in superposition - its not that it tells you something and we expect some differences from that in reality. [1] https://arxiv.org/abs/2203.01982 https://arxiv.org/abs/2203.01982