I've read "White Fragility".
While I completely agree that the stories in this article are hugely problematic and represent issues that need to be solved, I think books like "White Fragility" are not helpful in solving them. This is due to a focus on group identity, and describing "White" as if it's a monolithic group of people, all with the same culture, emotions, and reactions.
Another interesting aspect I identified while reading the book was it's description of the emotions that one can expect to see when confronting white people about race issues: the description could have been used to describe any human being you will ever meet when you accuse/blame them for something that they did not personally do. It really does read like a horoscope in that sense.
I find it ironic that people on HN, who are typically super data driven, get on board with works like "White Fragility". Diangelo is one of many academics from the humanities departments who are incredibly pseudo-scientific. Data is incredibly scarce, measurements and studies even less so. Statistical knowledge isn't present in the vast majority of these folks. Typically, the "scientific method" is reading and writing essays/novels. When you don't attempt to quantify a problem, you can't propose solutions and then measure their results. You instead just keep yourself busy finding ever more ways to describe the water to the drowning person.