Now, this is not a useless heuristic, as the fact that older people have had more life experiences is definitely worth taking into account.

More relevant to the social reasons for the heuristic, they have also had more time to accrue power and allies. For most people that is what respect is about (awareness of their power to influence your outcomes conditional on how much deference you give them).

The heuristic can be very useful when respecting the older person is not really a matter of whether he/she is right or wrong, but more about appeasing power. It can be very useful to distinguish between the two situations.

Oh, yes, those were the two points I prepared in response to your first paragraph. You nailed both, exactly!

Signalling social deference and actually considering an opinion to be strong Bayesian evidence need not be the same thing.