Economics as a discipline can certainly be positive and scientific. Developing a model to predict the elasticity of demand, collecting the data, testing the mod

Economics as a discipline can certainly be positive and scientific. Developing a model to predict the elasticity of demand, collecting the data, testing the model, and independently replicating the results is certainly science, with all the pitfalls of doing science well. Answering whether one should raise/lower the price of some product P to maximize profit for company C is a normative question. BTW, calling economics a social science doesn't mean that economics gets to shun scientific methods, nor is economics any less a science than physics. Being a social science just means that the discipline will apply scientific methods to develop models of, collect data about, or directly experiment on human subjects. (Physicists, and normal humans alike, don't view individual particles as human and feel no qualms about performing whatever experiments they might feel like, no matter how high energy.) Humans tend to be understandably nervous about being test subjects and this has limited the pace-of and extent-of progress that economists have been able to make. The extent to which economics relies on inaccurate models ("rational consumers, competition, efficiency" as you say), is the extent to which personal morals, governments, and societies inhibit progress. One thing I find very cool about economics is that it can put bounds on how much these barriers to our scientific understanding cost and benefit us. There is a cost to having a better model than "rational consumers" and we can know how much that is. But whether or not one ought to pay that cost is outside the realm of economics. Overall, I think your claim is more rightly applied not to Economics the discipline but to Economists the practitioners.