For example, even though they may have performed many (no idea how many is "many") shuffles in testing, the conditions in the field are doubtless different. I find it entirely plausible that food, beverages, humidity, dust, and grimy fingers could impair the device's ability to perform a good shuffle. I have no evidence of this, I just find it plausible.
Further, these are mechanical devices that are not immune to failure. Couple this with the fact that "card room" and "maintenance" are two terms very rarely used in the same sentence, and the opportunity for badness to occur clearly exists.
I have absolutely no idea what SM's field testing program looks like. I'd like to think that they periodically have someone at customer sites testing that card room's actual playing cards in the worn machines to make sure that the shuffle is appropriately random. Heck, card rooms should take it upon themselves to do this sort of testing during their relatively quiet times. I'd be shocked if any do, though.
Even with their problems, hand shuffles are well understood, and when they go bad (which *can* happen), someone usually notices this and corrects this before too long. Even though I do not have any personal reason to expect the SM devices don't do a good job, I do believe that if they do go bad, this fact might not be noticed in your average card room for an uncomfortably long time.
I'm trying not to be paranoid about this, but I believe there are reasonable reasons to be suspicious.
-- Nick Christenson npc(deleted the rest)