It absolutely does make sense to talk about single factors in isolation - it is obvious that lowered rate of violent crime is a good thing.
It is also likely that trying to change society in order to lower this statistic will affect other aspects of society, and not all of these changes are wanted - I think we agree on that. Thus it becomes an optimization problem, where you try to maximise some set of values - typically safety from violence (no matter wether it comes from private persons or governement, against your person or your property), freedom (what we typically call "liberal values" i.e. the freedom from someone else telling you what you can and can't do), predictability (i.e. making it unlikely that the economy crashes tomorrow, or Putin invades, or tax law rapidly changes, etc.), economic growth (you can afford more stuff next year) and many more.
These values have to be weighted against each other, and it is by no means certain that the current setup is the ideal one or even the only good one. Given that the boundary conditions are always changing, what worked yesterday may be less optimal today, so we must continiously reevaluate both our goal function and how we try to maximize it. As an example, there are many places which are further from being a "police state" than the US, while still having relatively strict gun laws - this is only a insignificantly tiny part of the huge patchwork which is society.