My conclusion (1) is that in order to be able to predict a system behaviour you need to be able to stabilize it, that is you need to be able to control it or equivalently speaking to regulate it.
I cannot predict what you are going to do tomorrow. But If I can order you what you are going to do tomorrow, then there is a good chance I can also predict what you are going to do !

Bonus thought
Linear systems are very rare. Most systems are non-linear. But control engineers use and abuse of linear systems. The important reason is that systems can be locally linearized. Then, locally, you can apply linear control tools. You can stabilize your system and because it is stable you remain in the initial local place and you remain stable. That is all the beauty of it. So yes, somehow, linear systems are not as rare as they seem. Otherwise, there would be no aircraft autopilots because aircraft or spacecraft dynamics are not linear systems.
About Economics / Aerospace comparisons, you can also read this post:
2 lessons Economics should learn from Aerospace

(1) (not engaging Taleb, he does not seem much against regulation, for instance he wants companies to be prevented from becoming “too big to fail” but he is probably less regulation prone than I am. I am for instance for a safety authority on financial products like there is an aviation safety authority in charge of certifying airplanes (FAA).)