I stated that actually sharing your premises and assumptions is remarkably difficult, and therefore Hanson, Cowen, and Aumann were wrong in some ways. I have been reminded that they are also right in some ways.
I have to agree that disagreements aren't honest, but only because I know how they're not honest.
Most discussions aren't really about what the participants believe.
I mean that tons of very, very common beliefs have been labelled as socially unacceptable. Selfish beliefs. Beliefs about one's superiority, particular or general.
Other kinds of beliefs, such as particularly revealing beliefs; beliefs that make you vulnerable, according to the accepted status game rules, to insults or credibility trashing. Beliefs that make you emotionally vulnerable.
A third kind, slightly different, of beliefs that are too important to allow challenges to. If your lifestyle leans heavily on certain beliefs, and you cannot afford doubt on these beliefs, then it is rational not to make them vulnerable to criticism or questioning.
Taken together, the patterns that ties these beliefs together is that they're the kind you particularly care about, and that saying them out loud is embarrassing. I know, because I do all this too. These kinds of beliefs are usually avoided in conversation at all costs.
And, I should emphasize, this is usually the correct strategy, as I tried to imply above. While some of them are embarrassing due to social constructs, others are embarrassing due to actually being declarations against interest.
Arguing about things you don't really care about is utterly pointless, a complete waste of time. As far as I can tell, anyway, perhaps I've missed something. But I can guarantee nobody learns anything in these 'arguments.' If an argument isn't making me uncomfortable, if the prospect I might be wrong doesn't worry me, I know I'm not arguing because of the belief, and I require myself to have some other good reason, such as practising communication.
Part of the reason I hate abstraction and love concrete examples is that it is so common to cloak these serious, care-about beliefs in layers of abstraction. It sterilizes the argument, makes them safe. Even if you lose convincingly, you can tell yourself it was a loss at the abstraction layer, not on the concrete bits from which it was derived.
However, taken together, I can safely predict that what Hansen or Cowen call meta-rationals will never be noticeable until there's a tribe that gently encourages sharing of embarrassing beliefs. A tribe that forgives the embarrassing trait, that by contrast celebrates the courage necessary to deliberately embarrass yourself. A tribe that doesn't reinforce the aversion response and reiterate that you should be embarrassed about it.
Come to think, this makes me pretty angry. Furious, in fact. You see, many of these embarrassing beliefs are held by ~99.9etc% of the population. These social norms aren't showering contempt on epistemic mistakes, they're showering contempt on being human. Admitting them doesn't mean you're a bad person, they mean you're homo sapiens sapiens and I for one refuse to be embarrassed to be a human being.
 I find that it is part of my learning process to criticize what's said before I work out what they were trying to say. Noting and pinning down the errors in their actual words allows me to be comfortable with granting them anything they're right about.
P.S. Yeah I just used Hanson's razor on something written in part by Hanson. Being meta-rational is low status.