Instead of afflicting scientific research with a vague, culturally-dependent set of metaphysical expectations and constraints (naturalism), why not characterize the scientific method in a metaphysically neutral manner? I think a more productive alternative to “methodological naturalism” would be “methodological pragmatism”.
This would define science as a methodology for finding/creating models that work for various purposes without assuming or making any metaphysical judgements on the nature of reality. It would also not generate metaphysical expectations or confine interpretations of data to metaphysical narratives that are largely cultural.