That bit about literacy doesn't strike me as accurate. Granted, it's almost impossible to measure historical literacy so anyone debating it is dealing in very broad strokes, but literacy rates in western Europe are thought to have been pretty high by the 16th century - perhaps around 60 to 70% of men and 20 to 50% of women. I have a hard time believing that the population within the borders of the Roman Empire would match those percentages, or indeed that there are accurate ways to measure the literacy of such a huge swathe of the ancient world. (Remember, by "within Roman borders" we're talking about enslaved salt miners in Turkey, North African berber shepherds, Scottish borderlands raiders, Visigothic tribespeople and all sorts of other groups beyond the inhabitants of the city of Rome and the Italian peninsula.) Narrowing it down to "Roman citizens" would yield a very different and much higher literacy percentage, but that'd be cherry-picking.