Hmm, I've got to say I'm not really convinced. It seems like what he's actually talking about is science answering
factual questions, which can only be interpreted as answering moral questions if you've already arrived at some assumptions about morality via non-scientific, philosophical means. In particular, there is not a single thing he says in this talk that is genuinely a scientific answer to a genuinely moral question, unless you supplement the science with additional non-scientifically-validated axioms, which tell you why a particular empirical fact about the world, which is just a fact, should be interpreted using terms like "morality".
Science can, as he points out, constrain the possible philosophical positions one could consistently take, by providing facts that moral philosophies can't be inconsistent with (e.g. some kinds of utilitarianism might be ruled out, if they assume humans work in a way that, empirically, they don't). But that's somewhat different, and has been something people have discussed for a while (empirically studying morality was one of the things Nietzsche got famous for).