Isn't adding a confidence element overmodelling the standard? The confidence of an indicator can (and should i.m.o.) be specified in the methodological notes: how do you measure and what are the limits. The methodological notes can be referred to through the proposed document links. Than there is no need for yet another element.