Please let me know if my rationalization is amiss. =)
(1) All scientists have "beliefs" and "values" that might slant their interpretations
(2) Serious scientific papers are reviewed by other scientists before publishing
(3) Other scientists (reviewers) are likely to notice and object to biases they do not share
-------
(C) ANY slanted interpretations will generally have been removed (from the serios scientific papers)
(A) I don't think (A) contradicts premise (1) unless we assume that "beliefs" and "values" are "biases" (which could be a different gap in the argument that this necessary assumption question doesn't exploit). On another note, (A) says that the reviewers do not always have biases likely to slant their interpretations of the data in those papers. Let's negate (A) and say that reviewers do always have biases likely to slant their interpretations of the data. Does this negated answer destroy the argument? Not necessarily. The reviewers could have biases and yet be able to somehow, keep the biases from making it into publication. How? That's outside the scope of the argument. Perhaps the reviewers have an internal system to pass the documents amongst each other so as to check for biases they don't share with other reviewers. In any case, it is not necessary to assume (A).
(B) is necessary. If scientists share all biases, the reviewers will not notice and object to SOME biases. If this is the case, the conclusion cannot follow because no biases will be removed (all biases are shared).