Earlier quoted context omitted.
Maybe you read a different post, but the one linked here simply argues that "any black person will tell you, when black folks date online they don't go to OKcupid," then lists a bunch of sites targeting black users and concludes that the OKCupid data is not representative of the public at large, despite the complete absence of supporting data. I don't see an analysis of data, just a hypothesis.
Actually, it said "when black people who don't want to date non-black people date online, they go to one of the black-only dating sites rather than dating sites aimed at the general population." It did still simply provide a hypothesis, but it is at least a plausible hypothesis that might explain the data. One would need further data to determine its validity, of course.
it is at least a plausible hypothesis
It might seem plausible if you share the same preconceptions as Ta-Nehisi Coates. But where's the data? Plenty of things that sound plausible to some people turn out to be unsupported by data, particularly when they are issues that people are emotionally invested in and issues that, like this one, have heavy social and political components.Racial views in the US are absolutely filled with (and arguably dominated by) gross misconceptions that seem plausible to many people despite being untrue.