Indeed! The question is, "How do you tell?" The "how do neurons work" research has gotten down to the level where the decision-making seems to involve quantum phenomena where we can't take the lid off and peek inside. Theoretical physicists wonder if there are more than just four dimensions, but haven't nailed down anything concrete. We can sort of see back to the beginning of our universe, but not into anything that may have been before it or beside it or anything like that.
You can definitely say it's "not outside our universe," whatever it is, because containing absolutely all of everything is part of the definition of "universe." But that doesn't actually answer the question of how it works in any meaningful way either, merely gives a more optimistic outlook about the odds of us figuring out how to understand it.
Interestingly, there's at least one experiment I've heard of where they were using an FPGA and genetic algorithms to "evolve" circuits, and it turned out to not be practical at the time, but analysis of the resulting circuits found them working in strange and unusual ways. Some of which seemed to be delving into the same kinds of quantum phenomena that we now know neurons use. So, at the very least, if there is some deeper layer it doesn't seem like it's a "protein chauvinist."
There are indeed multiple ways it could work. And it may be tough to decide how to draw any boundaries. Is it some totally separate realm that only interacts with ours in the one area? Or is it something that's simply a little outside of the four dimensions we can normally perceive and it's tied in everywhere in subtle ways and our cognition is merely the only spot where we easily notice it? We might try to model it in a number of different ways depending on exactly what we find. But we're almost certainly going to have problems trying to fully understand something so different from what our brains are built to work with, especially when getting into things like this where it's definitely possible that attempting to understand it could have a feedback loop with how we think at a fundamental level.
And I can't really think of a way that we could know for sure if our universe is deterministic or not. You'd have to be able to see multiple runs of it and observe if they were identical or not... It's kind of like the theoretically O(1) "randomize data, check if sorted, if not then destroy the universe" sort algorithm...
Well, the question is whether our thoughts are deterministic or not. If you reset the universe to the same point multiple times, would everyone necessarily do exactly the same things? Or might there be variation? There being an extra-universal influence on our thoughts that wouldn't get reset gives the possibility of non-determinism, even if there is some ability to predict what it might do in known circumstances.
Actually running that test though would be... difficult. We only get to see one of the runs, so we have nothing to compare to.
So is our sense of free will an illusion? Or meta-information that's leaking in from somewhere due to incomplete sandboxing? Really hard to know for sure. But, at the same time, does it actually matter?
Thing is, there are quite a few questions about our universe which simply cannot be definitively answered using only information from within our universe.
Take "free will" for example. Does our thinking arise entirely from natural phenomenon, or is there some extra-universal component to it? Well, if it is the latter, then the only way for us to find out from inside the universe is if the universe is built in a way to make it obvious. If there's some discontinuity between cause and effect with regard to thinking or similar.
But if there is a supernatural influence on our thinking, why the heck would it be bound by our limited perception of the time dimension? It's outside every normal dimension except time? The characters in a novel, if they start trying to examine whether or not they have "free will", find exactly what the author wishes them to. And it will be as perfectly consistent as the author cares to make it. You can only break out of the sandbox if there's a flaw in the sandbox, deliberate or otherwise.
"Simple models" are good. Using the simplest model that produces accurate predictions reduces mistakes and confusion. But the model is not the universe, and the apparent simplicity could be an entirely local phenomenon.
Why? If the answer is "no" then applying a proper punishment causes the nebulous whatsit in charge of the person's free will to change their future behaviour.
If the answer is "yes" then applying a proper punishment adjusts the programming of their brain in a way that will change their future behaviour.
The only way a "yes" makes it harder to justify punishing someone is if you overexpand a lack of "free will" to imply "incapable of learning".
As far as we know, there has been not one single violation of conservation of momentum from the uttermost dawn of time up until now.
And because we know that, any unusual reports that would seem to imply such a violation may have happened are obviously false... Ties up the loose ends.
The chemical stuff could be explained by alterations to thermal expansion. Less expansion would cause less pressure, and spiking pressure is a critical part of getting an actual detonation. Would also reduce the amount of wind though, so the climate would possibly change substantially.
Electronic stuff failing is rather more difficult to figure out without wrecking people's brains, compasses, etc. He probably should have left that alone and just let the electronics fade away since without gas expansion generating electricity to run them would be impractically expensive.
It may well be a "tightly-laced reality". It's just not this one. Perhaps the answer to a match not working in the world the hero is transported to is that the fundamental chemistry of the universe is different and our protagonist's body has obviously been modified to match. Or else the difference is some specific alteration where human metabolism can still work, and yet phosphorous can't generate a high enough temperature to ignite cellulose. The fact that he still has a match after transportation to such a different world where probably only his mental pattern is actually making the jump is the harder part to explain.
Similarly it might be possible to create a world where firearms and engines don't work by changing how much effect temperature has on the expansion of gasses without wrecking other things too terribly much.
But... we're talking about fantasy, not hard sci-fi... It's about the people, not the specifics of the physics of the universe.
If you interpret it strictly, an answer of "yes" puts you in the space of "I used to beat my wife, but I have stopped." An answer of "no" puts you in the ambiguous space of "Either I used to beat her, and I still do, or I never have and therefore can't have stopped."
The question is which of those two possibilities people will assume. Which will depend on the context and what they already think of both you and the person asking.
Placing it on an empirical foundation would be an enormously difficult task, but fortunately it's not particularly necessary since, like geometry, you can put it on an a priori foundation stemming from some basic observations about human nature.
Human beings tend to prioritize according to some simple, general rules, and natural selection ensures that those few who throw too big a curveball don't propagate. So you can take those rules and extrapolate them into a description of how a group of human beings will react to various economic pressures.
"Man, Economy, and State", By Rothbard, is probably the best and most complete analysis to date.
To answer the original example question, the primary role of "underconsumption" in the Great Depression was that politicians could use it as a boogey-man to justify seizing control of the money supply and the manufacturing and agricultural sectors of the economy and running them in a manner to further their own interests at the expense of the general population. "The People's Pottage" (terrible title, I know) by Garett Garrett has a good, blow-by-blow description of the sequence in which this was done and how each takeover was used to manufacture justification for the next.