To be clear, there was a review policy prior to the new one, but it's focus was rather different, the primary concern was on not revealing proprietary information. Which is a fairly reasonable restriction, though I would be interested to know what, if any, papers were asked to be withdrawn or retracted under the old policy.