Editorial Anonymous has, for writers, some good news and some good news about children's books reviews. The good news, she (?) says, is that good reviews can help sell books. And the other good news is that bad reviews won't hurt selling books.
I have a more nuanced opinion. More and more children's books are review-proof: good or bad, reviews won't make much difference to series franchises, celebrity books, brand-name authors or merchandise. All of those depend on marketing and saturation. Where reviews matter is in public libraries and schools (which themselves serve as a staging post for wider readership).
Good reviews do still matter to this institutional market, and bad reviews (or no reviews) have both a primary and secondary effect. Middling or worse reviews for an author without a built-in audience mean that not only will librarians be more likely to give the book a miss, but its publisher will be less inclined to fork out more money for advertising and promotion. As the legendary Mimi Kayden said, "one or two stars won't do it anymore."
And lets not forget the ALA awards, which consistently provide a bigger boost to sales than any other award out there, save perhaps the Bluebonnet. If The Graveyard Book had been published to indifferent reviews, it would most likely have not won the Newbery Medal. Not because the award committee members are slaves to reviews (although I have seen reviews used to kill a book's chances), but because the members and the reviewers are the same people. Sometimes literally, but more pervasively in the way they imbibe the same historical tradition and, however shifting, "standards." While the Newbery Medal only gilds the success of Gaiman's book, it was essential to the shelf-life of, say, The Higher Power of Lucky (although maybe that's not the best example of my point, as the book got respectful but not overwhelmingly enthusiastic reviews prior to the award).
Certainly, reviews mattered more when most juvenile hardcover was destined for the institutional markets. But certain books still need success there (if only there, often) to allow the author the go-ahead to publish the next one.