Not having read the book from which this mini-sequence stems from, I raise here three points hoping they won't overlap with some future post.

The first one pertains to this quote:

As long as information is ”walled off”, many, many contradictions can be maintained within one head.

Strictly, this is not true. That is, having separated modules for different informations is surely a sufficient condition to make the brain be able to hold many contradictory informations, but it's not necessary: a trivial counter-example is a database holding in different lines different statements about some fact. A more poignant proof is the callosotomy, showing that two modules continue to exist even if there are no more connections between them. However, the presence of contradictory information by itself is evidence of the modularity only under the unlikely assumption that every module tries to achieve internal consistency.

The second one regards the connectivity of the whole "brain graph" (if the modules are mapped to vertices and accessibility relationships as edges): while a complete connectedness seems highly unlikely, it is appealing to think to the brain as a strongly connected graph, i.e. a graph in which there's a path from every node to every other nodes.

Third, we must not forget that these modules are a 'software' or 'cognitive' reduction of the brain. Evidence from neurofeedback, or simply the possibility to control blinking frequency, points to the creation/destruction of non-innate connection between separated modules. It would be fun if you could learn, through neurofeedback, to send false anticipatory brain activity for the wrist movement.

So there are many modules in your brain, some of them conscious, some of them not.

This to me seems plain wrong. I would say that none of them is conscious, otherwise you have just moved and fractioned the problem. But maybe I'm misinterpreting here, and you really meant "some of them produce consciousness and others don't".