There is an extensive amount of research in the area of peripheral perception and subconscious acquisition of information, including tests where information is flashed on a screen for durations shorter than what the conscious mind can perceive, or from distances further than what is resolvable by the conscious mind (such research extends from the 1890s with Boris Sidis up until the present day). The research findings have shown a general trend to "priming" - the information is somehow absorbed and does lead to modification of behavior based on that information. For example, in one experiment a list of words was flashed briefly on a screen - far faster than could be consciously perceived, and when a second list was displayed and the subjects asked to choose words on the screen, there was a definite correlation between the words chosen and the words that had been previously flashed on the screen - a "priming" of the subject had occured.
I agree with the argument that was cited - you cannot "read" in the PhotoFocus state. Yes, the words are blurred and concious comprehension is probably not going to occur. However, perhaps what is occuring is that there is enough information perceived by the eyes, coupled with the brain's working knowledge of the shape of letters, that even with the letters blurred the brain is able to make "best guess" matches on what each of the letters are and thereby "absorb the correct words". The brain is a master pattern making instrument and, given scant bits of information, will create patterns that are meaningful to it. Given the context (a book that is being read), training of the brain in reading(PhotoReading is designed for ages 13 and up - part of that might include the necessary training of the brain in understanding sentence structure and letter shape recognition), and pattern recognition skills of the brain I would believe that there would be little or no problem with the brain properly decoding information that is on the printed page that is being PhotoRead. Also remember - when you're PhotoReading the text is generally only about 10-20 degrees off of the main line of sight, not something like 60 degrees. I would imagine that for fine processing of detail and information the more off-center the text is the more difficult it would be for the brain to fill in missing information and thereby create the correct gestalt.
With all of the above said, I am probably wrong. I don't know diddly about neurophysiology, psychology, or the structure of the eye.
