That's dumb. How can a test be predictive if it's weighted to support a particular outcome?
With regard to the weights of the SAT that produce a total score: there is no good argument for any weight, and - logically - no good argument against any weight. In such a case as this “the rule of insufficient reason” says apply equal weights.
Also: the component scores are made available, along with the combined score. So, if anybody wanted, they could apply a different weighting scheme.