Regardless of your view of the strength or weakness of circumstantial evidence it has nothing to do with Methodological Naturalism and science by definition.Yes circumstantial can be very strong especially when all other chances of other events are zero or that there is none for any alternative. Circumstantial evidence is very useful in court, because, you make the assumption both sides will have it, and if one side does not, even though it's not a certainty, the jury can rule in favor of the one with it and against the one who lacks it. This is why "reasonable doubt" is explained to juries. So they know they don't have been proven beyond doubt, but reasonable doubt, means circumstantial evidence is useful.
It applies only to the nature of some kinds of evidence in Law. Circumstantial evidence may be useful in the Court of Law.
Please document the use of circumstantial evidence in Methodological Naturalism. Hint: There are no references.