Yeah, i could be wrong about the level of awareness, i am a datapoint of one.
The performance part is interesting, but almost irrelevant imo. If the results had been that abusing your teammates improves their performance, it would still be wrong to do it.
I worry the people causing this problem are more likely to take the “abuse == tilt” information and use it to justify their behaviour :(.