Personally, I would not consider "Locality" to be a morph. Locality differences in all species of all animals are simply pockets of individual populations that through segregation from outside populations have tended to display a number of similar (and oftentimes) polygenic traits. While the locality traits are obviously heritable they are NOT the result of a mutation to one (or several) of the genes that occur naturally in the population in question.
I would describe a morph to be the result of a heritable mutation to one or more genes that determine the color or pattern (or, in some cases, both color and pattern) in an otherwise wild type individual. Taking it one step further, I would also elaborate on the "heritable" aspect of this definition by stating that the mutation should only affect the appearance of the animal and not it's overall physical form (an example being: two-heads, no tongue, no eyes, blindness in general). In other words, it should not affect the animal's ability to grow to maturity and mate.
Obviously, this would not mean that albinos (for instance) are excluded because the trait makes camouflage nearly impossible for babies and it could be argued that it decreases their life expectancy in the wild. What I am talking about is simply excluding mutations that result in more than a change in the animal's color and/or pattern.
With this (admittedly long-winded) definition in mind, I would not consider animals that have been selectively bred to bring out one or more traits to be morphs since there is no real mutation going on. Selective breeding is more akin to "steering" or "nudging" the path of evolution in a general direction whereas morphs are mutations. Breeding morphs is less like a "nudge" to evolution and more like a full on tackle to force the result of the breeding.
(No offense to Jeff's Pastel "morph" or anyone else that has selectively bred boas, pythons or any other animal to create a new "look". I'm speaking of a definition that works outside of the reptile community and general marketing practices. It is generally accepted that certain selectively bred traits are considered morphs within our reptile culture. This works well for us because it follows the practice of keeping things simple. The main difference between a selectively bred "morph" and a "true" morph is that selectively bred morphs are "created" by man whereas true morphs are mutations that are generally "discovered" by man. You could selectively breed until Doomsday and never create an Albino - at least not without the Albino mutation.)
Combination morphs would also not be considered morphs under this definition. They would be viewed simply for what they are: a combination of two or more morphs phenotypically displayed in one animal.
So, long (completely PERSONAL) definition, made short, a "Morph" is defined as having all of the following characteristics:
-Is the result of a MUTATION to a gene (or genes) controlling color, pattern or both in an animal
-Is heritable
-Must be able to be separated from other genes (in other words, bred into or out of a given population
-Cannot physically alter the animal's general form, whether or not the alteration is considered beneficial or detrimental
Yes, there are some morphs that can be lethal in their homozygous form (or, more rarely, when combined with other morphs) but these tend to be the exception rather than the rule. In their heterozygous state they still produce a viable, fertile animal that could reasonably be expected to live to maturity.
There you have it, I'm always long-winded, but I think (and unfortunately for some of my friends and family) and speak the same way I write. I like to cover all of my bases and define things as accurately as possible. Sure, I may have left something out, if so, PLEASE, let me know. There are probably some aspects of this that I hadn't considered.
-----
.
.
.
.
Stay United!

I'm still not sure if it's weird that my best friend is a two year old boa named Ronin. He's quiet, non-judgemental and listens... what more could you want?