Indeed, "rational actors" is intrinsic to neoclassical economics but not Austrian economics, as neoclassical economics at least pretends to be a scientific theory, which requires the ability to make predictions. Neoclassical economics has as its foundations, its "standard model", human psychology, and thus requires assumptions about human psychology that must be obeyed consistently in order to form mathematical models at all, i.e. the "rational actor" who follows the exact descriptions of how people should behave according to the models.
Austrian economics on the other hand agrees with the foundations of human psychology but largely dismisses that these can be mathematically modeled. Austrian economics is thus not a scientific theory but in some sense sort of an anti-theory, it argues against having any scientific modeling at all.
Sure, Austrian economics has "upheld the test of time," because it's unfalsifiable. It makes the claim that the economy can't be predicted, and thus it both predicts everything, and also predicts nothing. It is impossible to make any falsifiable predictions with an Austrian economic model because it makes no predictions at all, other than predicting that the economy can do whatever it wants whenever it wants.
The moment you admit the economy can be modeled, it becomes unavoidable to concede that some sort of interventions can be used effectively in it. This is why even Keynesians concede to some minor level of planning can be useful. Austrians rely on not conceding in the first place that the economy can be modeled, and thus any intervention into the markets becomes inherently irrational.
Your point about "rational as a result of human reason" is also just word salad because every economy ever in the history of mankind is the result of human reason. Humans cannot construct anything without reasoning about it. It's not an interesting point. The point is not that "the economy is the result of human reason" because this is redundant, the point for the Austrians is that the economy cannot actually be modeled at all on a macroeconomic scale, that human reason applied on a macroeconomic scale thus must always be based on faulty, irrational ideas, and thus will only serve to hinder microeconomic decision making.
My point of view is that the Austrians are correct that trying to mathematically model the microeconomic human psycholgy is absurd and the "rational actor" notion in neoclassical economics is nonsense, but Austrians are wrong to then conclude this makes macroeconomic understandings inherently infeasible. We can instead base macroeconomic models off of objective and measurable limitations that the markets must operate within, real-world physical limitations that restrain decision making. A macroeconomic model would only predict the boundaries of which the economy could physically operate, with a probability distribution of how it would operate within those boundaries. Austrians are wrong to conclude that because you cannot have a "rational actor" that is predictable and mathematically modelable, that therefore the economy is indistinguishable from randomness. It is not, because on a macroeconomic scale, physical limitations in decision making will cause predictable trends to emerge. But on a microeconomic scale, yes, you can hardly predict anything.
Marxists do not pretend to have a reductionist scientific theory of the economy like neoclassical economics does, with its own "standard model" that can predict everything on every level from "first principles", which, ironically, neoclassical economics, which claims to predict everything, can't actually predict anything.
Marxism instead reserves its measurable and falsifiable claims only to the macroeconomic realm. Indeed, the Austrians are correct to say that there is no mathematical model you could construct to predict exactly the price that a specific item will trade for in a very specific place in a very specific time. But they are wrong to also believe that you could not identify and predict long-term macroeconomic trends for prices, which those prices would then, on an individual scale, vary around that trend with a probability distribution.
Austrians are correct to dismiss the premises of neoclassical economics but then go an even more absurd direction instead.
Marxists reject trying to build a reductionist model for the social sciences. Any reductionist model that truly reduces the economy down to first principles would be indistinguishable from particle physics. It is nonsense to suggest that the "first principles" of the economy is anything else. The selection of "first principles" in the neoclassical economic "standard model" is arbitrary and without basis, and is not even supported by modern psychological evidence.
In the social sciences, due to the complexity, and the inability to truly reduce the theory down to first principles (because, again, if you did reduce it down to first principles, it would be indinstiguishable from the Standard Model of particle physics), you can only operate off of an overdeterministic theory, which dismisses first principles and instead only focuses on aggregrate trends and emergent properties.
