Those things are all perfectly indictable, though - and my indictments do not stem from a failure to attain perfection; but rather, because of the way their real world implementations behave, they are prevented from anything but the complication of society's ability to achieve their noble goals.
I don't intend to say that coercive systems are abusive because they contain an element of coercion by design. Rather, the people they (aim to) abuse (by design) in practice diverge significantly and often hypocritically from the malefactors whose intolerability was so great they caused the hunt in the first place.
I am not simply arguing coercion is empirically unethical, nor that coercion must be perfect, nor that there are no differences between different instances of coercion.
A difficult example where you need to coerce is when someone has Alzheimer's. If your mother has Alzheimer's, you had better provide care for her because the very material (therefore identified because of reliable differentiability of physical matter) brain disease (the opposite of an immaterial, and only socially-identified disease) that prevents her from accessing her existing body of evidence of the world can still be supplanted for her sake, by you as her child, so the kind of capacity of 'self-honesty' that Alzheimer's steals from the elderly can be compensated in a way where you take responsibility to 'donate' her self-identity back to her.
A simpler example is: if you are crossing the street with your young child, and your child darts out into the street in the path of an oncoming car, your duty is nonetheless to forcefully grab them by the arm if possible, because the child wants to survive instead of die by mistake.
I guess that I am trying to say that exactly because each of your examples (courts, law, taxes, mil) fails to meet the exceptional ideal they establish to allow their exception, they just tend towards abuse of privilege (in the sense that the expectation of the injustice from using the privilege is greater than the expectation of the injustice resulting in not using the privilege of coercion).
The thing they have in common is overestimating the frequency of hard successes that their proponents intend. They often inflict more damage to 'false positive' peripheral individuals than they prevent by coercing the 'true positive' individuals they are meant to stop. Therefore they stray very far from their aim. It is the aim that society is improved - but still respects the individual - that necessitates justifying the coercion that transcends the danger of catastrophically legislating coercion.
- courts are currently so beyond their capacity that the norm in >90% of cases (iirc the number correctly, but i am unsure that the statistic is accurate; merely that the percentage is unbelievably high suffices) is to state a lie over one's guilt and accept a plea bargain because the courts are so crowded due to draconian law that they cannot serve as instruments of justice without causing greater, indiscriminate injustice; the pressure for courts to move forward hastily + the laws whose inappropriate scope and severity harm the ability to achieve justice means their combination results in a failure to meet their social obligation to their optimal capacity of ensuring justice for all. You really have to believe there is a benefit to the guilty for living in a society where punishments for crime must be optimally minimal in misallocation. Justice is not the most ruthless form of litigation that results from encouraging self-inflicted punishment a priori for people who become accused. This is more self-referentially crucial to the ideal of justice itself, irrespective of the drug war at all.
- the military often grows because of the work of people who would prefer less of their taxes go to killing and more towards American prosperity. An example of this is Trump's "America First" doctrine having the opposite cost of his more important doctrine of devoting a plurality of funds to national security. The contradiction between the military and taxes is not because the military kills people that Americans don't want nor that taxes don't improve expenditure towards things America wants, but rather because they achieve unity in spite of expression as contradicting ideals:
+ A) our national security is the most important thing we can spend money on, because survival comes before the ability to even benefit from any other expenditure, and
+ B) we only permit taxation in the first place because taxation gives a non-zero-sum benefit for fulfilment of society as a collective beyond each individual only worrying about their own security
My argument is itself a heuristic argument, I would say:
expectation of iatrogenics (harm of the healing) of avoiding iatrogenics of epistemically-flawed institutions < expectation of iatrogenics of epistemically-validated but non-institutionally-remediated concerns < expectation of severe unknown iatrogenics you become exposed to by succumbing to compulsive institutional remediation of lesser, known iatrogenics
If I understand correctly, all you need to do to heuristically estimate which is most harmful in the big picture is to pick a large number K that you use to compare probabilities [0]. The severe outcome for which the probability P(Outcome | Outcome > Harm severity K) is most harmful for those probabilities where P(outcome | outcome > k) is proportional to k*outcome; so large, widespread minor policies can incur massive damage.
[0] Taleb, Silent Risk (forthcoming), Chapter 1