I think that there is a line to be drawn, but not with respect to the degree of the effect, but to the semantic concepts that the white = good, black = bad apply to.
For example, I have black eyes and hair, but I never once felt I was less a person because of this abstract generalization. I don't recall ever having made such a connection until when I started thinking about arguing against the reasoning behind banning blacklist.
Similarly, obsidian and black granite are typically considered beautiful. When I choose the color of my phone or display or car, I don't think white = good or black = bad. When there is context, this abstract association does not come to mind; rather, for example, I think that white can distract from the screen of a phone, or that it wears its stains easily. Or when I choose a dark theme for my editor and apps for ease of viewing in low light. Or with clothes; I don't think (black = bad) when I wear my black clothing.
So this is why I disagree with this; we associate ourselves with black things and make color judgments all the time, but rarely does the value judgment of white = good, black = bad figure in these judgments (modulo actual reasons in particular contexts, e.g. stains easily, vs. hard to see in low light). An argument that this "psycholinguistic effect" actually happens also needs to address the many other categories in which we make color judgments in our lives. (Question: does a strongly racist person significantly and subconsciously disassociate themselves with black products more than the average person? Beyond conscious decision on color to symbolize their attitude, and the subconscious generalization from this symbolic desire.)
With respect to blacklist, it seems to me that this is the most unfortunate product about the whole discussion. I believe that hardly anyone has associated the word "blacklist" with derogatory attitudes against blacks, but now someone has made "blacklist" and black of race so tightly bound semantically that we can't help but think of race when we now see the word blacklist --- even when nothing close is ever being intended by the people who use the word.
I think that my model that there are many semantic categories of blackness for which the value judgment (white = good, black = bad) doesn't figure at all as a "psycholinguistic effect" is more accurate. If this is indeed the case, then ironically the whole discussion over "blacklist" has caused more harm than good, since it is enlarging the collection of meanings that remind people of derogatory attitudes against blacks.