What you seem to be proposing is that this trait is somehow quantitative (like skin color or height in humans) and sex-linked. The more of the PB alleles you get, the more phenotypic expression you see.
Therefore a male could be PB because he could have two copies of the PB allele while a female could only have one copy. However, if it was that simple, a het male would look just like a PB female. Yet, as I understand it, they don't look the same.
Furthermore, most quantitative traits like that have an additive effect (the more of the particular allele - the more trait expression). But the PB gene doesn't seem to result in the addition of anything, rather it seems to result in the failure of one of the pigment genes (similar to amelanism). I don't see how a faulty allele could result in an additive or codominant effect. Generally when an allele is faulty, it is recessive the normal allele, because the normal allele provides enough of the trait to overcome recessive allele's contribution.
The other question would be about het males. I assume that male snakes have some sort of dosage compensation mechanism which would shut down one of the Z chromosomes in each cell. If a male PB was het for the Z linked gene, wouldn't he show variegation (like a calico cat) where some of the body showed the normal pattern and other areas showed the PB expression?
I just don't think we can eliminate the possibility of it being some sort of sex-influenced trait rather than a sex-linked trait.
There are three male and two female phenotypes. I see three explanations for this:
1. This could be explained by a sex linked, incompletely dominant gene (with ZZ, Zz, zz, ZW, and zW phenotypes), but this only works if there is no dosage compensation in male snakes.
2. It is an additive sex-linked trait. However, in this scenario, somehow the PB allele would seem to have to add something to the phenotype, not subtract something, as I mentioned above. I think the PB allele is subtracting something from the normal phenotypic color expression. And again, you need to consider how dosage compensation would affect this in het males.
3. It is an autosomal sex-influenced trait (like pattern baldness). The phenotypic outcome of the trait is regulated by levels of hormones (male hormones in this case - this is based on the assumption that female snakes, like female mammals, produce some male hormones), and so the trait appears co-dominant in one sex and recessive in the other. Therefore the phenotypes would be
male AA = normal
female AA = normal
male Aa = het (pale male)
female Aa = normal female
male aa = PB male
female aa = pale female
Fortunately, this hypothesis would be easy to disprove. You would have to examine the phenotype of females with 100% KNOWN genotypes (not assumed het by their color pattern). If this scenario is true, the known het female should look just like the normals.
Of course, it could be much more complicated. There could be multiple genes, one sex-linked and one not which have an epistatic interaction, etc.
-----
Chris Harrison