I totally agree that the reviewers should read author responses and take them into account in their evaluation (otherwise, what's the point). I don't think the reviewers need interact more with the authors after that point (though they should with the other reviewers and AC)