'Electronically' created white noise is indeed extremely effective to calm down a fussy or crying infant. It also often works instantly, making it certainly a welcome and powerful tool in the arsenal of the sleep deprived parent.
That said, I found its instant effectiveness to be somewhat concerning, as it almost seems to flip a 'brain switch' in your baby - quite different from the longer and more laborious procedure of shushing it to sleep the old fashioned way (which likely works on the same principle, but involves lower volume, more frequency variation and, last but not least, actual parent-infant interaction). Bit of a no free lunch situation.
Driven by concerned temptation, I did some digging, but ultimately couldn't find any large scale studies on the safety of exposure to white noise in human infants - there is one study in the rat model though that suggest delayed auditory cortex development in baby rats that were exposed to white noise 24/7. (http://www.hhmi.org/news/white-noise-delays-auditory-organiz...) - which btw is exactly what sensorimotor contigency theory would predict. (http://nivea.psycho.univ-paris5.fr/Synthese/MyinFinal.html)
While I don't think the majority of SNOO users will operate the device this way, I would imagine the effectiveness and convenience of the device may mean that there are quite a few babies who will spend a lot of time in the SNOO, approaching long term exposure to white noise.
I am not suggesting that this will cause long term problems to the infant's auditory development as is the case in the animal model, but it is something to be mindful of.