Almost nobody assumes absolutely perfect information; even Adam Smith was relying on approximations of actual behaviour. Most economics relies on the assumption that information is close enough to perfect that we can just assume away any difference.

Behavioural economics has proven unreliable, with results that are (at least) difficult to replicate, and non-predictive.

Sure, you can say behAvioral economics experiments are unreliable. But even if this field did not exist, assuming close to perfect information is simply naive in my opinion. Unless you account for insider trading, which the pessimist in me believes is rampant. In truth, none of us know, but I think it’s worthwhile to consider the alternative to what is taught in textbooks.

Not an economist, but I don't see how, given that I'm specifically mentioning people's expectations about the future, not just cold hard facts. Those expectations may or may not be rational, but they're still driven (in part) by news.

That doesn't assume perfect information in any sense