The goal is to publish solid advances in the state of the art, game-changing work, research that pushes the boundaries of IR. We are not looking for incremental improvements.William Webber wrote a great post about the rapidly rising field of complaining about aspects of SIGIR. Highly entertaining and worth reading.
Jeremy Pickens posted the reviews of his rejected paper and a discussion of them.
If you want to join the discussion, Ian started a Google Group, SIGIR Meta.