Current IQ science is persuasive on its own, even in respect to race. Past "race science" is not persuasive. Those who view the former as a continuation of the latter are betting that it's more probable that liars have gotten well practiced than that the hypotheses are sound.
There's no reason to think popular opinion, even of today, is determined by its correspondence to truth rather than by its utility, and much reason to think the opposite. Therefore, if mainstream scientific consensus converges on popular opinion, it's a bundle of lies.