I'm neither for this nor against it. I think that questions that I would have would be around its limitations, because everything man made has limitations. Do cards ever get damaged due to jams as a couple of people have wondered? Have the test uses suggested some sort of percentage of time that it typically happens? For example, maybe for a large operation, 1 in 10,000 is labor saving enough to be acceptable. Does it somehow sense the difference between cards with thick card stock and thinner cards? When it says it sorts cards, does it really sort cards like card number one is dropped into this stack, two in the next etc., or does is just know that this card is different from the others so I'll drop it in a stack that hasn't been used yet? Are there conditions that the operator is expected to adhere to, such as all the cards need to be face up or face down and all of the cards for any given run should be from the same set or have the same thickness?
I'm sure that I would be a real pain in the neck to a salesman, but these are the kind of things that I would want to know.
I'm sure they wouldn't tell anyone, but I would just be curious if it works by OCRing the card and finding card number (essentially reading the card) or does it know the card is different from other cards by picking out, say about 32 random spots on the card image and storing the color of that location for comparison to other cards in the same sorting run? In other words, is it reading the card to put them in order or is it just sorting them into like piles? No matter what, I would think that it can't actually have a thousand stacks to drop cards into, so what does it do when it exceeds the limit of stacks of cards to drop stuff on to? Also, what is the limit?