My take is that there is plenty of room for, and research into, a range of topics in economics including those which reexamine certain assumptions and which challenge the concept of market infallibility. However that work is rarely incorporated into the dominant narrative of the profession, which quickly reverts back to a simplistic Econ 101 worldview.
I think George Akerlof, 2001 Nobel Laureate, provides a good illustration of this. He was rewarded for his work on issues of asymmetric information, most notably his paper "The Market for Lemons." Michael Spence and Joseph Stiglitz shared the prize.
The Lemons paper explored what would happen when, in a market, sellers knew more about the product they were selling (whether it was a good quality product or a bad one) than the buyers did. Ostensibly a model of the used car market, the equilibrium outcome is that only "lemons" are actually sold. The basic reasoning is that as buyers are uncertain about the quality of the car they are buying, they're not willing to pay full value for quality used cars. Why pay $10 grand for a car which might actually only be worth $5 grand? Since offered prices are too low, potential sellers of quality used cars exit the market and only the crappy cars remain, which are sold at an appropriate crappy car price.
But this of course wasn't really a paper about the used car market, it was a paper about asymmetric information using the used car market as a handy illustration to make a much more general point about a world where asymmetric information was commonplace.
So where was George Akerlof wrong? It's because he thought:
[I]f this paper was correct, economics would be different.
The paper was correct, its insights and reasoning are taught in undergraduate classes, and the paper itself is taught in graduate classes. Scholars do, at times, apply asymmetric information models to other areas of research. But it didn't make economics different, at least not in the way I think he means.