> because there are no consequences or accountability when they're wrong
I did see the parents' suggestion and I addressed it thought not explicitly. Blackstone's ratio is often quoted as:
It is better that ten guilty persons escape than that one innocent suffer.
But it isn't about the ratio, it is about failure engineering. Which is about building a system in such a way that when it fails, it fails in a specific way[0]. The point I'm making here is that while the proposal sounds nice (and like Blackstone, let's abstract "triple"), it too has consequences and failure that must be accounted for. Which the parent explicitly mentions about being overly cautious.
The issue at hand here is that in law, like in many real world applications, there is no global optima. Even in a highly conditionalized optimization function that has temporal dependence, our world and morals rapidly change and our true function (under the naive assumption that it exists) is intractable.
Therefore we must account for failure. In our case we need to determine the consequences of how we respond to when {LEO,judiciaries,politicians,etc} make mistakes.
I am not arguing that we shouldn't refine the system. Nor am I arguing that we shouldn't hold these entities accountable. What I am pointing out is that there are consequences to whatever we choose to do and we need to be extremely careful with trying to understand those consequences. I'm sure harsh consequences could result in something akin to a naive interpretation of Blackstone's ratio because it would require substantial burden of proof, but the reason we would meet Blackstone's is because we would prosecute almost nothing. There's an obvious consequence to this kind of policy: any crime of sufficient complexity (or containing sufficient obscurity) would never be prosecuted. And almost certainly this would result in many crimes not even being investigated because there would be an understanding that the likelihood of the crime leading to actual prosecution is very low and due to the resources being finite. I don't think you need to be very intelligent to see how such a system would allow the wealthy to avoid prosecution in all but the most egregious of cases.
This is not to say that we shouldn't find a way to right our wrongs. That those unjustly/mistakenly prosecuted should not receive some form of compensation to right the wrongs (even if the wrong is fully reasonable in-situ. I would argue that ensuring we better compensate the wrongly convicted would put pressure on even individuals to use more caution. But I think we need to be careful and forward thinking to figure out the details here and I'm nowhere qualified to make an even reasonable guess.
It's important to remember that the reason most of the problems in the world that exist are not due to malice or due to a cabal trying to dominate the world but rather because the world is exceptionally complex and there's an asymmetry in the ease of doing what is wrong compared to doing what is right[1]. But there is a saving grace, in that this all lives on a spectrum, and pareto efficiency exists (while it might take a lot/infinite of resources to "do good" it does not to do "pretty good").
[0] A prime example is skyscrapers. You design them to fall in on themselves so that if they do fall, they do not take out neighboring towers and create exponentially more damage. But this is a key process in any engineering design and in textbooks they will often discuss bridges.
[1] A simple example might be when you realize you are about to miss your exit and too quickly merge one lane over to make your exit not realizing that there is a person you couldn't see that had to slam on their breaks. The result of which is that this can create phantom traffic experienced long after the event. You won't hear screeching, crashes, or any of that. It is the result of a cascading chain of people having to change lanes and brake to respond to this and the negative feedback loop of others making similar "mistakes" in response to trying to avoid an accident themselves. Or maybe you didn't miss your exit but are just trying to dodge an object on the road that you had no capacity to see and the truck it fell off of had no way to know, because the strap holding it in place broke because the manufacturer who made the part had a mouse get into the machine which happened because one of the workers that day had a PB&J. The point is that there's a long chain of events that in no way could you know about or reasonably consider which all need to "go right" for the final event in a chain to also "go right." They do not require malice nor even stupidity, but just life/random events. It does not mean we should spend all our time over analyzing the details and complexity, but it does mean we should be aware that this complexity exists, avoid rash decisions we we have the ability to, and to consider that there are faults and consequences to no matter what we propose (and thus we shouldn't get get upset when our ideas get pushback nor should we abandon ideas just because there are ill consequences). We can't be perfect but we can (always) do better.