I think Egapal and Littleindiangirl covered it.
I would vote that if spider is somehow proven homozygous lethal (very hard to prove something by its absence) that would seem to me to qualify as co-dominant due to the homozygous being different.
Also, a book that talks about the co-dominant and dominant form of the same mutation would be very unfortunate. I see enough of that terminology thrown around that I don't doubt that it's written down somewhere. As already stated, the mutation type (recessive, co-dominant, dominant) is defined by the relationship of the phenotypes (appearance) between the different genotypes (heterozygous and homozygous mutant) so it doesn't change. The pastel mutation is still a co-dominant mutation type regardless of if you are looking at a pastel or a super pastel, the difference is that the pastel is the heterozygous genotype and the super pastel is homozygous for the pastel mutation.
Cool that we finally have a ball python dominant mutation type example in pinstripe where the heterozygous pins and the homozygous pinstripe are reported to look the same (same phenotype) to fit the dominant definition. And you can see that it's the genotype that's important to know what the babies will be like with only the homozygous pinstripe producing 100% pinstripes.