If the majority of the abusers fall into class X and target class Y, and your goal is to explain or stop this behaviour, then that is informationally relevant, yeah.

This does not imply that all or indeed a majority of all members of class X are abusers. (And more importantly, this article doesn't go anywhere near making that claim.)

Not all victims of abuse are black or women, but black women get a disproportionate amount of abuse. Should be ignore that, too? How do we hope to provide systemic solutions if we ignore such important data?

That sounds an awful lot like profiling to me, and the social narrative is that profiling is wrong and immoral. Or is it ok to profile as long as you are profiling someone you perceive to be a member of a privileged class? I have a hard time keeping the rules straight.