I, too, have greatly enjoyed this debate. I certainly believe there are rules to CM, but I think some of the dispute has been how one interprets "rule". Some see them as regulations (inviolate) while others see them as guidelines and heuristics, and therefore evolutionary and dynamic. Some see them as analytically derived, while others see them as empirically discovered, . Still others see them as dictum while others see them as social/cultural norms that emerge organically within the social group.
The closest analogy I can think of is the work of the architect Chris Alexander. He wrote 3 books in the 70s on recurring "patterns" in successful (attractive, harmonious, enjoyable to live within) buildings and built environments (One being A Timeless Way of Building - especially for Tira). These patterns were distilled from his study of building over centuries and formed a pattern language through which people could communicate complex ideas in terse terms. An example is 4 story limit - limit the height of buildings in a residential community to 4 stories. His ideas are not widely adopted in architecture, but they were picked up with gusto in the computer science community as they were struggling with the design and architecture of massively complex systems and sought heuristics to help guide the process.
The upshot of all this is that I see the rules of dressings as heuristics collectively learned through experience over time which then emerge as norms within a given group. As in many complex systems, some of these rules can be in tension in a given situation, and priorities come into effect and thus some rules get broken, but it is only the expert who knows when it is the right time to break a given rule. He understands the consequences of the break, because he knows (if only subconsciously) the rules that he is breaking. Hence, it is the realm of the expert to break rules. Novices need first to learn and apply them.