> the laws … tend to be fundamentally authoritarian and unsound in principle.

Let's take the (big) step of presuming that that statement is true. We could further agree that changing the laws would be best — both to prevent future injustice and feasibly to relieve ongoing harms. That would be far more democratic and acceptable to society. The pardon power could also undo ongoing harms where legislative changes are impossible, if it were used across the board. Leave aside the queries of whether doing so would violate separation of powers by acting as a form of super legislation or super veto; and would violate the "faithfully execute" clause. The singleton use of the pardon of such offenses, only in a case of quid pro quo, does not remedy the posited unjust laws.

> But the broader perspective is the US government is irredeemably corrupt … and there is some level of bad behaviour that just needs to be tolerated.

In brief: this reduces to whataboutism (a form of the fallacy tu quoque); the fallacy that additional evil remedies the original evil; and presumes the validity of the conclusion as a premise en route to justifying it.

> The pardon power is, almost by definition, only exercised to corrupt and bypass what people agreed was an appropriate legal process.

Its proper use is as a retrospective safety valve to account for human fallibility in predicting the future and setting down prospective legislation. I think we would need to see some data to claim that it is only (or even predominantly) used to "corrupt" the legislative intent instead of remediating its imperfections.

> I'd be pretty surprised if past pardons were consistently above board; given the low-risk nature of a corruptly offered pardon I'd assume it is pretty routine.

This is mostly speculation, not argument. And corrupt pardons, especially at this scale and in large quantities (take the January 6 pardons, although many would say those were not so much corrupt in the monetary sense as corrupt in the sense of condoning those who helped try to grant someone power outside the system's rules) are not low-risk at all, from the considerations of adherence to the rule of law and institutionalism / systemic integrity.

> If I look through Wikipedia I see pardons that look superficially similar to Zhao's in other presidencies. Financial crime, sentence served out in the past, president intervenes to pardon them. Why did the president take an interest? Probably not because they were outraged at the moral implications of some banker being mistreated by the justice system.

If so, perhaps we should be able to find an example or twenty of quid pro quo after the passage of so much time. If not, that lack of evidence would undercut the posited defense substantially. Other possibilities are that the guilty somehow redeemed themselves through recompense and further good behavior. Regardless, the posited defense is largely speculation again, and reduces to the same fallacies noted above.