Devils advocate here but your argument is that they should change their safety policy because the job has great safety statistics? What if they change tactics and danger goes up?

This would probably be both expected and preferable. The way they currently act often puts the public at huge risk in exchange for officer safety. Given the police are trained professionals who chose their own occupation any added risk should fall on them.

This is a statistical question that can be checked by looking at other countries. The data is exceptionally clear that per-capita the US sees an order of magnitude more police killings than the majority of the world [1] [2] (side note, don't make Brazilian police mad) but I couldn't find an easy source for comparing per-capita police mortality rates due to violence. They are honestly pretty low in the US but the quick, un-cited, articles I found discussing it said the US rates were higher than most countries. This implies that not only do we kill more people with our policing tactics, but those tactics are leading to even higher policing deaths. So, loose loose. But, again, I don't have good sources for this. Everyone tracks police killings. Not many track police deaths because, I suspect, it isn't a dramatic number in reality.

[1] https://en.wikipedia.org/wiki/List_of_countries_with_annual_...

[2] https://worldpopulationreview.com/country-rankings/police-ki...

Do we take into account the people they handle in those safety statistics? I'm pretty sure we don't.

A guy got shot after being swatted not so long ago and that would be a data point with no injuries for officers, masking the real victims.

Then that danger is going to have to be balanced against the reduced danger to the general public

That's not Devil's advocate, that's just foolishness. First of all their job is "protect and serve" not threaten people with summary execution. Second, their instant escalation does nothing to keep them safe. If anything it makes them less safe.