Yes, this is the primary theoretical difference between Kolmogorov and Shannon. Shannon only deals with "zero-order" information, whereas Kolmogorov information deals with "n-order" information.
Using a fixed fifth-order structure for his theory is only slightly less wrong than using a fixed zero-order structure (i.e. Shannon). From what you are saying, he is trying to reinvent Kolmogorov, badly. It is puzzling that someone who nominally knows so much about information theory is apparently unaware of Kolmogorov's redefinition of the field, since it is really the essence of the matter.
What is even more puzzling is how it is that someone who nominally knows so much about information theory manages to have NO papers published in the relevant technical journals to that field, but instead publishes a book on the topic.
Or perhaps I should say it is temporarily puzzling, that is until one becomes aware that technical journal articles must undergo a nominally rigorous peer-review process before the are accepted for publication, while book manuscripts do not.