>randomness might not be so random when it comes to human agency

This is where the “delusion” comes in, imo. Ultimately, it seems many believers of free will also disbelieve in the laws of physics. You seem to think humans can defy randomness through sheer willpower.

I agree that the sense of voluntariness is significant for individuals. My critique of compatibilism is only that compatibilists seem unclear on the limits of their domain. It seems that this pro-free will argument applies to the feelings of an individual or perhaps for the convenience of everyday conversation, but falls apart when we consider communities or societies. The feeling that “Jon could have acted differently” is different than saying “Germany could have acted differently”. The more people that are involved, the more they can be modelled as statistically determined rather than agents.

This is because ultimately, we are beings without libertarian free will, or “delusional” free will as you call it. And the farther away we move from our individual frame of reference, the more clear that becomes. I feel like I can choose whether or not to commit a crime, but I know for certain that some percentage of the population will “choose” to commit crimes today. It would be a mistake to assume a whole society could shift based on collective spontaneous individual choice, it is statistically impossible. This does have implications for the way we ought to correctly talk about nations or about widespread social issues (e.g. why do the poor choose to be lazy?).

I disagree with S. Harris about many things but one thing I’ll agree is incompatiblism forces you to take judicial reform seriously. I feel compatibilists skirt around confronting the full implications of having no libertarian free will.