I tested mine, and it does happen as noted. But it seems to be somewhat of a red herring. You must have the lens cap on and the viewfinder blocked very well to detect it. I put a piece of dark blue masking tape over my viewfinder, and it still let in 1000 X as much light as thru the top LCD, so I had to really completely block it.
With the lens cap off at very high ISO, in a dark room where I could barely read the exposure, shining a bright light on the top LCD did not change the exposure. I guess if I were to take 10 minute exposures with the lens cap on, a light hitting the top lcd panel might impact exposure time. Of course, in the event a light was on the top LCD wile doing a 10 minute exposure would ruin the exposure in any event. Its hard to imagine how the almost undetectable negligable leak could impact me, or any real world image.
I'm waiting for a expert to tell us what real world impact, if any, it might have.
