In my previous post on underdetermined questions, the distinction between an underdetermined and a non-underdetermined question was somewhat vague and not sufficiently defined.
I suggest that a good metric for underdetermination is as follows. An underdetermined question is one for which there is no evidence sufficient to create consensus. If most people agree on a single answer to a question, then it is not underdetermined. If there is significant division of opinion, then it is underdetermined.