I mean, we see the exact same uncertainty all the time in real-world applications of the hard sciences as well.
Weather prediction, for example, is still just as inexact a science as macroeconomics in its application, even though it's entirely physics-driven and we have a pretty complete understanding of each of the variables involved on a theoretical level. The system is just complex and chaotic enough that understanding the theory of weather doesn't mean that we can successfully model the real-world behavior of weather.
This doesn't mean that we should conclude that all theory relating to weather is incorrect - that would we throwing the baby out with the bathwater - it just means we still have a ways to go when modeling the real-world complexities that come with the theory.