Bias itself is not a flaw in an argument. Bias may tell us where to look for flaws in the argument.
Not discard, simply distrust. Remember all those Microsoft studies comparing the total cost of ownership of Windows and Linux? Finding flaws in the studies was hard. Discovering an a priori reason to distrust the studies was not.
Simply put, any article like this is going to make claims: "X is a fact." When the source is biased, and X confirms the bias, it's much harder to trust that X is true.
People who believe in the equality of human beings typically believe in it strongly enough that they find it difficult to objectively analyze putative evidence to the contrary -- they're damn stubborn. Meanwhile, people who are racial supremacists of any kind in this day and age have self-selected themselves as pretty headstrong, stubborn individuals themselves.
And so it goes with The New Yorker: when you are deeply emotionally committed to a position, it's virtually impossible to trust anything you say. I admire Malcolm Gladwell, and enjoy The New Yorker, but really, could this article have gone any other way?