Is the acceptance ratio of enumerative Gibbs really always 1?

[axch]

Certainly it is if all the drg nodes are deterministic or principal (and the likes of #462 and #470 are fixed), but what about likelihood-free consequences or brush? Or down-stream (uncollapsed?) AAA nodes? This question is not prompted by any problems detected to date, but we do not have an accessible, recorded proof, and I, for one, mistrust my intuition without going through a formal exercise.