I'm pleased to see this new, updated summary of Tononi's theory of the conscious brain, Version 3.0. Lately I've published a piece with Gerald Edelman suggesting that our two approaches, "Neural Darwinism" and "Global Workspace theory," can be viewed as members of the same superset of "localist-globalist" theories of the conscious brain. I don't know if the authors would agree, but my inclination is to believe that we are all approaching the same great mountain of the conscious brain from different perspectives.
This article by Oizumi et al (2014) has several distinctive features.
1. The theory is presented as a formalism, which has the advantage of suggesting a common feature set that might allow a careful comparison to other "local-global" theories of the conscious brain.
2. New assessment methods of cortico-thalamic connectivity and information flow make such theories increasingly testable, notably the development of TMS-EEG in cortex.
3. A possible drawback is the relative rigidity of a formal theory in the face of more than 20,000 PubMed articles that cite the keyword "consciousness" or "conscious brain." There is much we do not know yet.
4. I would be curious about the relationship between IIT 3.0 and our own recent publication of "dynamic Global Workspace theory," which is also developing to deal with the great body of new evidence and significant theoretical proposals. (Baars, Franklin, Ramsoy, 2013).
5. There are by now more than 20 major consensual features of conscious brain activity compared to unconscious, and I would be very interested to see how IIT 3.0 accounts, for example, for a greatly improved understanding of the electrical activity of the cortico-thalamic system. (See articles by Anil Seth, Baars, and David Edelman).
6. I am also curious about the relationship between Tononi's phi and measures of Mutual Information, which seem more suitable to an "integration and broadcasting" conception of conscious contents. Mathematically phi and MI are linked ideas.
7. It is not clear to me how the authors would represent unconscious processes taking place in the thalamo-cortical core, e.g., work done by Goodale and Milner on unconscious spatial "maps" in parietal cortex, involving both allocentric and egocentric maps of body space.
8. The modern study of the conscious brain is only a few decades old, but it has moved surprisingly quickly. A number of other research group are now pursuing the same goal. Our goal should be to develop empirically testable predictions, including ones that can differentiate between different major approaches. Because we are only beginning to understand the nature of information encoding in the thalamo-cortical system, for example the roles of oscillatory signaling between cortical regions and layers, we will need an ongoing dialogue between emerging evidence and theory.