That kind of thing is absolutely routine and police unions are almost certainly the single greatest obstacle to improved policing in the US. But the way things are right now the police unions are tremendously powerful, and so reform efforts have to cooperate with them and find common interests. Shortening the cycle time from an alleged violation happening to discipline being imposed or not is one of those common interests. Whatever my feelings might be, complaining about the unions isn't really conducive to the type of work I'm trying to do. I'm a political appointee (so I have and express my own opinions), but I serve as just one part of a broader system that requires cooperation between parties. I can only speak to that system (one of many different efforts towards more just policing), but it happens to be the one that is producing programs like the one the article covers.
Police reform is inherently political, and so you ought to take everything with a grain of salt. But it's important to understand the context: direct feedback is nothing new and basically everyone (including the unions) agree that we need more of it. At the same time, the impact that direct feedback will have is limited because performance and discipline is not generally "data-based" in any meaningful sense and the unions (to some extent reasonably given the valid concerns around automation and employment issues) will generally be opposed to efforts to change that. So the bad news is that this program is not some silver bullet that will fix policing, just one of many small steps.
On the flip side, the reason that I commented on much of this discussion being based on misconceptions is because of the huge number of comments pointing out edge case issues or potential perverse incentives. It's vanishingly unlikely in my mind that any of these will materialize because of the critical context that I'm trying to share: first that police behavior is very resistant to any change at all and there are powerful institutional forces against things like increased leniency as a result of this program (although it's not at all clear that this is really a good thing), and second that the downsides of this type of program in terms of bias and unreliable information are already fairly well understood because the program is not really new but an expansion of a decades old effort.
The irony is that I am not intending to say that this program is good because police departments are effective, although maybe that's how I came across. I intend to say that this program is good because police departments are extremely resistant to change and have no problem throwing out data that runs counter to their interests, so this type of data collection can only really help to validate concerns from oversight and professional standards personnel.