I was (re)reading Weiskrantz' 2009, "Blindsight", trying to make sense of the different areas that were involved in this peculiar case of vision. I had tried to do the same earlier with Prosopagnosia 
[too many articles to mention, but it started with Gross, a graduate student of (who else?) Weiskrantz, Gross et al, 1972 "Visual properties of neurons in inferotemporal cortex of the macaque", and certainly did not end with the objections of the Tarr group against the specificity of this ailment, Gauthier, Behrmann&Tarr, 1999, "Can face recognition really be dissociated from object recognition?". The debate is still alive and kicking: Richler et al, 2012, "Holistic Processing Predicts Face Recognition".]
also to no avail. My frustration had almost reached a boiling point when I realized that it really did not matter where those phenomena are situated in the brain. Even if I believed in computational modules (or even computational neurons), which I most certainly do not, then I would still only need a functional analysis of those modules, and not necessarily their neural mappings.
My thoughts until now had led me to the following principles:

1) All neurons are equal (and none is "more equal" than the other!).
2) There are are no neural codes hidden within these neurons.
3) Neuronal connections are changed by experience.

Let us see it those principles can help me make sense of the issues mentioned.

Prosopagnosia
a) The fact that a neuron reacts to a face, does not mean it is a "visual" neuron, or that the brain area where it is located is a visual area. It could be connected to a group of non-visual neurons that are linked to specific visual neurons, those referring to the visual experience of a face. Vision is not an isolated process.
b) There is a difference between seeing a face, remembering a face, and recognizing it. Only the first is a visual process. Patients very often state that they do not see a face, even though they can distinguish certain features like hair, teeth, or external markers like clothing. Except for social awkward moments, they are able to function (almost) normally and even write books about it! (Oliver Sacks, "The Man Who Mistook his Wife for a Hat").
c) The only way there could be an intrinsic difference between faces and other objects would be if we "processed" them differently. Which brings us to the idea of specialized modules or neurons, and in last instance, to neural coding. Visual neurons would somehow have to signal that they are receiving face related stimuli to activate a specific face module. This is in contradiction with the principles stated above.
d) Either the principles are wrong, and neurons can not only detect specific features (beyond organ specific sensitivities), but also signal their presence to other neurons; or the principles are correct and prosopagnosia, in all its forms, is a memory disorder, or at least, not a visual deficiency. 
e) The debates should concentrate on the mother of all issues: Neural Coding, Myth or Reality? Instead of being sidetracked by the question whether one kind of neural coding (Zenon Pylyshyn, "Seeing and Visualizing", 2003) is more plausible than any other kind (Treisman&Gelade "A Feature-Integration Theory of Attention", 1980). That would at least make it necessary for everyone to explicitly state their assumptions and beliefs concerning the brain. (Which Pylyshin certainly does, even if he never offers any empirical evidence for the computations he assumes the brain is performing all the time. Certainly not in the form of chemical or electrical properties of neurons).