When you talk about signal versus noise, you are referring to measurable material effects. Whether a material effect exists is not dependent on whether someone notices it. When a signal is lost in the noise, it's no longer measurable, but the material effect continues to propagate. If someone spits in the ocean daily, no weather station on Earth could detect the change, but if the person is getting their water from a source not usually in the water cycle, then yes, the average volume of the ocean would increase by an incredibly small amount.
There are market trends that are large enough to be observable despite the chaotic nature of the system. But our collective ability to observe and comprehend the effects depends on how well surveyed the global economy is, the resources put into analyzing the data, and the ideology used by the analysts. If the signal was truly destroyed when it was overtaken by noise, then when corporations see an unfavorable trend coming they could raise the minimum signal to noise ratio to protect themselves materially by firing the analysts, erasing the data, and enforcing a more favorable ideology. This is the magical thinking. Companies and governments attempt this all the time, but it doesn't change their economic situation.
You've thoroughly laid out how the noise amplifies through each layer of the system. The problem is that the system you've described is far too noisy to function if noise wipes out signal. At the opposite end of the chain from the consumer, it's all noise, there's no signal left. Even when something like closing the strait is clearly big enough that it should have some effect, the farm managers have no coherent way of determining what the effect actually is. They're hiring and firing children based on random number generators.
Given that the system operates at least somewhat rationally, that can't be the case. "Signal" and "noise" are subjective concepts. One person's noise is someone else's signal. At each layer, the managers watch the decisions made by their neighbors for signals, make their own decisions, and then pass those on to others as noise. Then they wait for signals to come back in the other direction.
If you fix your perspective to either the consumer or the farm manager, they have no way of ever perceiving each other through their personal noise. But if you look at the system as a whole, their actions do impact each other, and there is an intact chain of decision makers who will continue to communicate the effects of each one to the other. This is evidenced in the physical product repeatedly transported from one to the other. For each step of processing and transportation, there has to be a person maintaining the link between them, or else the product would not arrive.
You have a convenient definition of causality. We can be done here.