See this paper for more info: https://arxiv.org/abs/2211.06738

Then we present our main open problem: is there a heuristic estimator that formalizes intuitively valid applications of the presumption of independence without also accepting spurious arguments?

The paper describes the concept of a "heuristic argument" and presents some work in this direction; however, it leaves as an open question whether a satisfactory formalization exists.

Resolves based on the opinion of Paul Christiano, Mark Xu, or other researchers familiar with the direction. Resolves N/A if there is significant controversy between researchers, and I deem myself unable to adjudicate with confidence.