I strongly urge that you have a really really good read of the technical data on in The PDF at the bottom of the page you linked to and then see if you can get someone from the company to explain the test standard they used to obtain the results they got.
Given the ability to create your own test standard you could come up with some very interesting results.
For instance If you stipulated that the test was to be done in a dark room held at 32 degrees but the fridge was to be encapsulated in its optional thermal blanket and was to be preloaded with a volume of 50% iced water in bottles with a test duration of one hour. That test would proove you have a fridge that has a 0 power consumption. (yes thats a massively skewed test but thats the idea)
Ideal test conditions are like statistical data, get enough of them skewed the way you want it skewed and you can prove anything you like.
