So far, all of the data I have collected was with a white balance card and not a gray card. I would prefer people to use a white as opposed to gray. There are variation between cards, but there are variations within a single card of a given picture. I acknowledge that this isn't a precise science (as calibrated photospectrometer would have been use otherwise), but I'm more interested in extracting general trends that tie in what I've learned about solar position, atmospheric effects, etc.