Jeff my problem with the color scale is that you'd apply it to captive produced animals too. IMHO this particular factor of phenotype is too easily manipulated which would render any relevance to wild populations meaningless. I agree that letting geneticists hand down sup-specific classifications is a bit disconcerting and I'm not sure if I buy into all the findings that are coming to light but contrasting these findings to a subjective data set doesn't seem to solve anything in my mind.
If I were to be interested in some such scheme I'd more likely follow pattern variations as I don't think these are as widely "bred for" or against. Again going back to coastals I could easily go alone with the idea of three distinctive regional "types" of temporalis based on pattern. Keeping in mind that I'm pulling these figures out of the air this would roughly follow:
Type I
Local: ground zero for coastals
Example: southern MD
Triad count = X
Lateral expansion of triads = .5
Presents of alternates = 25%
Type III
Local: southern MD
Example: Albemarle Peninsula
Triad count = .7X
Lateral expansion of triads = 1
Presents of alternates = near 0%
I would think that such a scale would more closely represent the spectrum of geographic phenotypes. Keep in mind that such a scale would not be totally conclusive. Historic gene flow would account for the presence of individuals within a given geographic area that are more typical of an ADJACENT area. This would follow your observations that St. May's coastals (southern MD) showed a wide range of phenotypic expression as Type I is adjacent to both other types, however, it would be rare to fine a Type III phenotype in the Jersey Pine Barrens and visa versa.
Not only would such a scale work for the classification of wild populations, but it would provide a framework for the captive production of geographic "types" without the problems of inbreeding depression locality breeding of rare snakes entails. By this I mean that genes from a Type I animal with Type II or III phenotypes could be migrated into those respective groups. This would not compromise the integrity (ability to breed true) of any group and would preserve heterozygousity in captive populations. It would also more accurately reflect the realities of nature be the elimination of the nearly complete barriers to gene flow between populations and sub specifics that locality breeding mandates.