And how many failed predictions have economists made? Even a broken clock can make some correct predictions. Moreover, there's no chance the economist's sophisticated math is needed for legitimate scientific purposes. When economists make correct predictions, they're using commonsense and inductions from history. The math gives them their epicycles to hide their failures and their propagandistic role in society. It also gives them the cachet and authority of rigorous science, so they can facilitate the "free market" frauds.
Of course societies tend not to perfectly implement the economist's abstract models. Those models are meant to be ceteris paribus, so external conditions would interfere with the system. What can make such a model nevertheless scientific is its ability to explain predicted probabilities, without relying just on commonsense and historical background knowledge. The neoclassical positing of Homo economicus isn't like that. This is plutocratic, social Darwinian propaganda hiding behind obscurantist math. It's a pseudoscientific disgrace.