The high prevalence of unintended consequences, delusional viewpoints, and
"bang-bang" control.
For example, let's consider the current story. Prosecutors got to this point of casual abuses of laws and ethics because there are no consequences. People cared more about the appearance of "being tough on crime" and expediting the prosecution of crime than on results, the unintended consequence is that now there are a bunch of out of control prosecutors who can harass people and organizations just because they feel like it.
There are plenty of delusional viewpoints such as ignoring or rationalizing law enforcement abuses, rationalizing that it's ok to jail people because of drugs or unpopular things said on the internet, or assuming that the government which commits blatant abuses in one area of intervention won't commit similar abuses in an area that the person wants intervention in.
And a near universal delusion of government intervention is that it is possible for an external group to know enough about a problem to effectively intervene. Commonly, there are profound knowledge gaps that aren't acknowledged and which greatly impair actual intervention attempts.
Finally, the real world bang-bang control problem is balancing a rod, like a broom, by moving the bottom end back and forth, The official version allows movement in one dimension and the only choices are to move left or right at constant speed. For cases where the rod isn't too far out of balance or moving too fast, you can keep it suspended indefinitely even in the presence of small, continuing, random disturbances of the system. You are always moving left or right. This uses more energy than a gradual approach that uses smaller adjustments when conditions are near balance.
Government intervention is not gradual or nuanced. And it's not uncommon to see government set up systems with fundamental flaws and then propose increasingly complex and bizarre fixes as the system goes more and more out of balance.
A classic example is the military-industrial complexes of the developed world, particularly, the US. The fundamental problem is that the system is a vehicle for transferring wealth from the public to a well-connected private network. National defense and security is lower priority whether by intent or not.
This has resulted in the sorry spectacle of a system that can keep track of the quality of the screws, but not the quality of the final product. There are plenty of major US defense contractors who have produced crap, overpriced products for decades (with really nice screws), yet still get plenty of business from the US government. There's no credible means to correct the problem because there are a small number of suppliers and any punishment would either impair the US's near future defensive capabilities or be trivial enough to ignore.
In summary, there's plenty of evidence that the approach doesn't work very well, due to the ignorance of the people controlling the system, the crudity of the means of control, and often just due to having terrible goals that are wildly incompatible with the needs of society.