Hmm, you're finally starting to write about things that I understand. It's about time chaos theory reared it's ugly head . . .
1. The description and classification of the network states is good, but does the introduction of a "strange attractor" concept add another dimension to this? There's a few pages in Artificial Life (can't remember the author, it's one of the ones I got you for Christmas last year) which detail evolution in terms of gradually increasing levels of chaos surrounding distinct areas of local optima. I don't have the book in front of me, but I seem to remember a chart depicting life on Earth, with the vertical axis representing disorder/entropy and the horizontal representing time. Over 5 billion years, the actual duration of time in which life exists on Earth is relatively small, however these periods seem to be extended beyond a linear increase in disorder (basically a "plateau" where the level of disorder stays fairly constant over a period of time). There was also a whole section devoted to genetic algorithms (GA's) that pointed out the same phenomenon.
2. The other concept which might expand your model, again related to the network, would have to be feedback. The model which you've outlined feels like a "first-in, first-out" (FIFO) scenario, where stimuli flow through the network components in a fairly orderly and progressive fashion (that's the way I read it, anyway). The introduction of a feedback concept could probably help to model situations where small "blips" in psychological state can trigger accelleration toward more manic sorts of behavior.
This might be a good fit with the strange attractor concept -- although the human brain isn't infinite, it's pretty damn big on a neuron-by-neuron basis. Use of a state model on such a complex network would imply virtually infinite variations in personalities and response patterns. However, people are actually fairly simple in comparison -- 5+ billion people all fitting into a finite number of personality "types." These personality types would represent the primary local optima, or the points that behavior tends to gravitate toward (the strange attractors). However, there are obviously other attractors in the pattern, those representing "deviant" behavior. Since these are less common, they would show as smaller, less "significant" areas of optima on a chaos chart. It's probably not quantifiable on the human behavior front, but I wonder if anyone in the engineering controls field has examined how differing levels of feedback tend to "push" a system state between periods of order.
By the way, I can also validate the point about the futility of having more than 4 people in a meeting. I absolutely despise large meetings -- there seems to be some sort of exponential relation between the number of people attending and the overall lack of productivity. Conference calls are even worse.
NOTE: I'm a proud father! I need to work on "attractors," "landscapes," and bias factors to keep up with my son.