But does it solve fluids or n-body gravity?

What of the/a dilatant fluid model of gravity which predicts the perihelion of Mercury does not jive with observation?

Doesn't that indicate that CFD is the viable way forward?

The point of the OP is that it's neat or maybe useful that semiclassical gravity solves NP-complete problems; but if the semiclassical model of gravity is insufficient to describe even gravity, why should it be sufficient to solve NPC problems, and what does a sufficient model of n-body fluidic gravity enable low-error predictions of?