You know, I would've said the same thing, but I think that isn't correct. Achondroplastic dwarfism in man has a homozygous lethal form, but it is considered dominant, not co-dominant. I think that other homozygous lethal traits are as well.
I think that your working definition (which is the same as the one in my head as well) is actually incorrect. Spider and achondroplasia are dominant because the mutant phenotype is dominant over the wild-type phenotype in the heterozygous form, regardless of what happens in the homozygous form (a dead/unformed versus indistinguishable homozygote).
In the case of the morphs we call co-dominant, which should really be incomplete dominant, there is actually a blending of the "pure" homozygous mutant and the pure wild-type form. Super pastels are the "pure" homozygous mutant, for example, and the regular pastel is a "blending" of that plus the wild-type. The reason we say we have to wait and find out if there's a "super" form or not before we label a mutation co-dom or dom is to find out whether a given heterozygote that we're looking at is actually the less-extreme, "blended" version of a "pure" homozygous mutant phenotype (which is invariably more extreme than the heterozygote in that case). Does that make sense? That was a run-on sentence and I might've just confused myself ...
I guess because you can't "blend" the lethal form (because it's a phenotype that essentially doesn't exist since it's incompatible with life) and the wild-type, you just consider a homozygous lethal mutation to be simple dominant ...![]()
(Co-dominant, BTW, is where BOTH traits are expressed equally ... AFAIK there aren't any traits in ball pythons that could be truly considered co-dominant, and most of the true co-dominant traits I know of are expressed on a microscopic level ...)
Honestly it's all just semantics. I think a lot of these terms are pretty out-dated when you talk to geneticists ...(They are wicked useful, tho ...)