[absolutelastone]

Observability means it can in principle be measured, not whether it is practical. And we can observe all the way down to the individual level where effects happen (prices), individuals set interest rates, etc. Aggregation is also in the intuitive column. At worst you could argue this reguires a subsequent step of logic, which I accounted for since humans do by design.

Thermodynamics is easy. I did it as a sophomore. It can be derives using statistical mechanics from classical mechanics but this is not necessary. The ideas of heat itself and its flow are ancient.

Speaking of cognitive biases, pointing out examples of errors does not provide a proof that something is worse than some unnamed alternative. Unless your argument is that we are worse at something that an omniscient and perfect version of humans. And if you go all the way back to my very first statement, I said common sense always works to some degree. I didn't say the predictions would be perfect. You are making the much-harder-to-defend argument that it disappears from value via some magical statistics effect you haven't identified.