TL;DR: Does it make any sense to judge the optimality of human cognitive functioning by comparing against unrealistic economic toy models?
It seems fashionable these days, to talk about how humans are not "rational" due to many cognitive biases. I think we need to think about this more carefully.
Take some complicated problem in game theory that economists cannot solve/explain today. How could one expect the participants of the game to be able to solve the game (i.e. play rationally), when the experts do not understand how to? So it is impossible for the participants to act "rationally". Simply put, thoroughly solving situations (even if theoretically possible) requires computational capabilities beyond access to humans. By the way, this includes most daily social interactions. So, it's clear that in many many situations, humans act out of heuristic rules rather than accurately calculating expected utility. This is what Daniel Kahnemann calls "Thinking fast and slow". Thinking fast (based on heuristics) is an optimal approach in its own way, where the brain's computational power is a scarce resource that ought to be conserved.
This brings me to my central thesis. When comparing homo sapiens to homo economicus in their handling of a toy problem or a sample situation, it might seem as if people have persistent cognitive biases. That is because they are using heuristics learned by nature/nurture -- heuristics based on millions of years of evolution of a culture for social interactions. It should not be very surprising that those heuristics seem to fail in the case of contrived and unrealistic toy problems.
In such a case, we could do one of two things: blame the humans for being irrational, or question whether "rationally" solving the toy problem has any application in actual human life.
By encouraging people to be “rational” we might actually be making everyone more stupid.
I posit that since human thinking has been honed (along with culture) to interact socially, there must be some underlying reason why humans consistently think this way. What that should tell us is that the toy problem being considered in isolation is probably deficient in modelling real life. So any "irrational" behaviour of homo sapiens tells us that the game cooked up for homo economicus to play is a bad model for the real world, and we should be careful about the conclusions we draw from that. A very interesting example of this is the Ultimatum game.
Economic models have their uses, and I shall not delve into that here. My point is that one should be careful about what we consider optimal behaviour, and therefore endorse and encourage. By encouraging people to be "rational" we might actually be making everyone more stupid.
In conclusion, I thought it interesting, what's encoded in the names Homo economicus and Homo sapiens: Economicus is defined by rationality, which is the ability to reason while, to be sapient means to be wise.