Dembski's source article speaks of "
lots of light reflecting in synaptic and nerve layers, but with regular patterns of empty holes with no scattering"; within the holes (actually, Müller cells), there is excellent transmission of light (with a funnel-shaped deally to increase the amount of light collected). They conservatively report that "
the müller cells can capture and transmit as much light as possible" (that is, given that the retina is back-to-front, this organization allows light through the maze of nerve cells and blood vessels. Dembski, though, crows that "now we find inside the eye optic fibers that transmit 100% through the layers of “bad” stuff in front of the cones and rods." Either he is carefully phrasing it so that he is speaking only of the transmission through the Müller cells--and thus his comment omits any admission that there is, in fact, loss of information through the synaptic and nerve layer scattering, or he simply misunderstands his source article.
The source article is far more interesting than Dembski's.
The most interesting thing on the Dembski page itself is
this. I must say, though, I don't quite understand the point of its presence there.