Oh, no ... now it's on PSG, too. I thought that maybe we would escape this.
Anyway, for those that are not familiar with this debate, here are, IMHO, two articles that do a reasonable job summarizing the discussion:
http://www.wired.com/2015/02/science-one-agrees-color-dress/
http://www.nytimes.com/interactive/2015/02/28/science/white-or-blue-dress.html
My objection to these articles, as well as almost everything else that has been written on this topic is that almost no one mentions the fact that the camera almost certainly was set to perform both auto-exposure and auto-white-balance. These can, and most likely, did introduce profound differences in how the colors and tonality were recorded in the file, completely independent from the physiological optics / cognitive aspects that most people have been invoking.
Specifically, most cameras AWB and AE systems would lighten up the dress and try to remove the blue cast by introducing its complement, yellow. Of course, most cameras' AWB and AE systems intentionally don't make "full" corrections, so they tend to leave an image in a somewhat ambiguous color and tonal state, somewhere between uncorrected and "fully corrected", and this is where the physiological / cognitive ambiguity aspect comes in.
Just my $0.02,
Tom M
PS - In case you missed it, another photo of the dress, but this time shown in context of its surroundings, is shown at the end of the NY Times (ie, 2nd) article I cited. Just so this image doesn't disappear if the link eventually gets broken, I'm attaching a copy to this post. From this, it seems clear that the actual colors of the dress, ie, independent of the illumination, camera automatic corrections, and cognitive ambiguity aspects is pretty clearly blue and black.