This is a question that's so loaded its disingenuous without context. You weren't asking me but here's what I think about it.
Do I think that police involved in bad shootings or another type of unjustified death have a really good chance of getting away with it? Of course.
Do I think that the scenario above of an unjustified police killing happens more to people of color? Probably, I haven't looked up the numbers though and I don't know where to find them.
Do I think the overwhelming majority of police killings are justified? Yes. But that doesn't mean it isn't a big deal when unjustified deaths go unpunished.
Do I think most white people in the US could get away with murdering a black person? No, and keep in mind this is closer to the question you actually asked than the question you wanted an answer to.
The original statement "It is true that white people kill black people with impunity in the US" is problematic without contextualizing it.
If he means in a general sense (which he most likely didn't) then its obviously a false statement. If I as a white man in the US murdered a person of color, most likely I would be arrested and convicted.
If he means in the literal sense that in a country of 350 million people that its possible for a white person to kill a black person and get away with it, well then you could probably say that about any demographic vs any other demographic because its impossible to make sure that never ever happens in a population that size.
What I suspect he really meant is that there's a problem with American police getting away with it when they are involved in unjustified killings of black people. But he chose to word it in an exaggerated and inflammatory way for emphasis. Then you also have to take into account that we were presented with that one sentence out of a larger blog post so perhaps he added the context that would have made that sentence make sense.
One of the failure modes I see often in today's discourse is that using a group term as a subject makes the associated verb profoundly semantically ambiguous. If I say:
"Brunettes like smooth jazz."
It can mean any of:
* There is at least one brunette who likes smooth jazz.
* Some brunettes like smooth jazz.
* Most brunettes like smooth jazz.
* All brunettes like smooth jazz.
* Brunettes are more likely to like smooth jazz than people with other hair colors.
* Brunettes are more likely to like smooth jazz than people as a whole.
* Liking smooth jazz is a defining characteristic of brunettes.
* Liking smooth jazz is a defining characteristic of people who identify themselves as "brunettes".
* Disliking smooth jazz is a defining characteristic of non-brunettes.
* Liking smooth jazz causes (some|most|all) people to dye their hair brown.
* Having brown hair causes (some|most|all) people to like smooth jazz.
When the group term has a long history of power imbalance (unlike brunettes for the most part) and when the verb has deep moral implications (like "murder"), then obviously these different interpretations connote wildly different things.
When you take that ambiguity and place it in the context of the Internet where context is stripped and nothing is known about the audience who will be interpreting it, you are setting yourself up for misinterpretation.
When you do that in a political environment where people are seeking power and stand to benefit from willful misinterpretation, you get, well, much of what US online culture looks like today.
There are obviously many deep systemic problems, but one technique to try to improve the quality of discourse is to simply avoid using groups as subjects in sentences. It almost never conveys anything that can't be better expressed in some other form.
Less so today than they did before BLM highlighted the way they repeatedly did so.
EDIT: I mean, two people actually think the recent spate of prosecutions of white law enforcement officers for unjustified killings of blacks is because in the last few years American law enforcement officers have just become violently racist in ways they previously weren't, and not a change in accountability resulting from public attention to the issue?
