Now please calculate: how many Julian years does it take to produce an error of 10 days? The answer is 1257 years. The question – at which date was the Julian calendar correct – can be calculated with the following amazing result (Illig 1991): 1582 – 1257 = 325Thus there are about 300 'phantom years' that history says exist, but didn't. Thus the whole period 600-900 A.D. never existed.
That's a very interesting theory for us medieval philosophers, who focus almost exclusively in the post Carolingian period, by reason of the paucity of literature in the intervening period. If that period simply did not exist, then the lack of literature can easily be explained, because no one existed to write it. If the dark ages did not exist, then Boethius, writing in about 520 (or 820 on the PTH theory) was nearly contemporary with John Scotus Eruigena (c. 815 – c. 877). There would be no 'gaps' to explain.
Other things are less conveniently explained (such as the venerable Bede, who appeared to have been writing bang in the middle of the phantom period). Presumably people counted the years as they went by. Assuming there were no written records whatsoever, and no archeology to speak of, and assuming that different people counted the years as they went by, with only the record of the current year. What is the probability of miscomputing the year? Indeed, what is the probability of different groups consistently miscomputing the date, so that they all all arrived at the same, but consistently wrong answer? How could they consistently lose more than 300 years?