Because, given the demonstrated ways racism "works" (as in, how it's practiced, and what effects it has when practiced between differentially enfranchised groups) it's not accurate?
> To insist that the "prejudice + power" argument is premised on bright-line "race == power" is to mischaracterize it to the point of being unrecognizable.
Would you mind clarifying the distinction between "differentially enfranchised groups" and "race == power" ?
By analogy, membership in a group that is more or less enfranchised than another could be seen as a Bayesian classification, while "race == power" is more Boolean.
One describes a tendency; the other asserts an identity relationship.
It doesn't map perfectly, but it's close enough for the bandwidth I have to keep contributing to this thread right now.