That's interesting!
I think that is a bit similar to me in ways, and probably why I don't lip/speech read (aside from having to be -uncomfortably- close)...
I only have a small 'window' of vision, and very near sighted, so having to be close makes that 'window' even 'smaller'. So I can miss a lot of things and not even realize it, even with visual text I have to scan over it and build a context, a lot of it going by memory... how much I use memory depends on how many times I am willing to scan over it and make sure I have it 'right'.
Interesting things can happen here, (you will know especially if you know about 'garden paths' and such in English, among other context 'tricks')
A fun example of this is "The car driven past the barn crashed."
We see "the barn crashed" but, the barn did not crash, the car crashed. This can be confusing. That particular phrase is small enough for me to catch the difference, but if it were longer... I might see something similar to the "barn crashing" effect, and say "what? what is this?!"
