There's a lot of speculative concepts being flung around in here.
I'm confused over how the concept of how trainability of the CG manifests itself in the body. The example given was with new runners finding it easier to run after 3 months. I'm not sure how this fits in the CG theory. It would certainly make sense for running to be easier as the body becomes more fit (build capilary beds, mitochondira, develop tissue strength, lactate buffering, etc.). Why a trained CG makes it seem easier isn't as understandable.
As an example, we don't inherently know how to ride a bike, we have to learn it. Once we learn how, we don't explicitly think 'turn left, turn right, lean left, lean right', this is relegated to a subconscious level and takes over the process. In this way, we've trained ourselves to ride a bike and riding itself becomes easier as we don't have to try to balance. This doesn't apply in running. We know how to run, that part is easy. What we lack as newbies is the conditioning, and it's the physical stress part that's difficult. But, as I said, this is easily explainable by improvement in physical conditioning.
The other part that isn't clear is trainability and how it the CG theory carries this to pace management. It's easy enough to understand that we practice pace management and train this way and can figure out this relationship between effort level and pace and distance. Again, like riding a bike, it may not be as explicit as thinking 'this effort means this pace' and this relationship is affected by a multitude of other factors. The way it seems be be presented by Dr. Noakes, the CG alters the perception of effort to enforce (or at least, recommmend) an output.
The problem here, it doesn't match the learnability concept.
Our cues to pace management and results rests on the effort levels we perceive and their relationship to pace. We learn how effort perception works out over the course of a given distance (environmental issues aside). It wouldn't make sense for a CG to alter those levels as this would alter our perception and our supposed learning against this cues has now been compromised.
Another example Dr. Noakes presented was the hypothetical situation of attempting to run a marathon in record time but consciously running the opening 10k at personal 10k PR pace, compromising the overall effort. Dr. Noakes posits the CG intercedes 'enough is enough' and forces the runner to slow down after 10k, then sets out an 'acceptible' pace for the remaining 32k. There's far too much speculation in here and an inconsistent presentation of how the CG works. Why in the initial 10k is the runner allowed to ignore the CG, yet forced to follow it after this? There's no doubt the runner will be exhausted after that first 10k, but it's highly speculative that they would be repaced to an 'acceptible' level. It's also nearly carte-blanche in defining 'acceptible' in this situation. I have a feeling any result would end up being considered a demonstration of the CG in action.
I don't have a problem in accepting that our perceptions of effort can change and this generally depends on the situation. I personally have never found it 'easier' to run harder at the end of the race (and have often lacked this supposed end-race guarantee the CG theoretically offers up, I've shuffled over the final k or 2 many times). Just the opposite, this has been the hardest part, although I can do it because I know the pain won't last. In fact, I think most of my PRs have me fading at the end.
In conjunction with this, the only part of us that can perceive effort/stress/pain is our brain but it can only present it against what's there...the stress inducing factors almost always exist before they can be perceived (outside of delusion).
I also don't have a problem regarding trainability and it being part of both our conscious and subconscious minds. My reference to this was in terms of experience and learning, and in a much earlier response to 'The Light' I indicated the learning works at many levels in our minds, not just the top 'conscious' level. I consider this something already known and don't see the need to apply a CG wrapper around it.
What I don't agree with the anticipatory aspects Dr. Noakes describes as being part of the CG, and these seem to be the primary aspects he wishes to center on. I see the CG as being reactive against physiological state rather then proactively anticipating a state at a later time.
But much of this depends on how large a wrapper Dr. Noakes wants to make the CG. The more known aspects of behaviour and physiology it attempts to wrap, the less relevent I consider it. It ends up becoming just a catch all.