When I refer to "perfectly rational" beings, this is a different concept with different implications than if I was referring to a consensus formed by numbers alone. In my thought experiment, I conceive of perfectly rational beings as theoretical beings who only affirm true propositions, hence the name "perfectly rational". If you imagined the domain of true propositions, this domain would be identical to the domain of true propositions affirmed by perfectly rational beings. Here, the difference between consensus and perfectly rational beings becomes more clear. A consensus can be wrong. We might say that the political consensus about a specific candidate was wrong. However, perfectly rational beings cannot be wrong by definition, as they only affirm true propositions.
I don't want to throw out too many ideas at once, so I'll just start with this and clarify the rest of my statement as the discussion requires it.