Actually if you go by the definition of heterozygous we are incorrectly limiting it in the ball python community.
Heterozygous really means having an unmatched pair of whatever genes you are talking about.
Heterozygous doesn't require any particular relationship between homozygous normal, heterozygous, and homozygous mutant appearances. That relationship is what determines the mutation type relative to wild type (recessive, co-dominant, or completely dominant). The genotype is only depending on if the pair (one from each parent) match or not.
So, not only is it correct to speak of a heterozygous albino (one albino mutant copy and one normal for albino copy of the gene pair at the albino locus), and also a heterozygous for the pastel gene (a visual het), it's even correct to talk about a het for a completely dominant gene like spider or pinstripe might turn out to be. If a spider phenotype animal has one copy of the spider mutant version of the gene at the spider locus paired with a normal copy of the same gene it has a heterozygous genotype of the spider locus.
By realizing which co-dominant and perhaps some day dominant snakes are heterozygous you can use the same rules to predict their breeding results that you learned with recessive. For example, het X het = 25% chance homozygous, even if the hets are pastels and the homozygous are super pastels.