Actually, it might be confusing at first, but blueapplepaste is doing the square the way it normally is done. That’s not to say that using “n” wouldn’t work. I understand why it seems easier. If the normal allele that pairs with a mutant pastel allele is normal, why not just use “n?” It’s not standard, but it can be done if it doesn’t confuse you. We would usually use P for normal allele opposite a pastel (p) mutant allele because that’s just the standard punnett method.
What you need to remember is that each bee has a possibility of throwing (we’ll use n) sp, sn, pn, or nn. That gives you a possibility of 16 outcomes in a 4 x 4 square. At this point I would not bank on anything that results in ss. In blueapplepaste’s example, a ppsS would, as you said, be a super pastel het spider. Since spider is considered in the herp world as dominant, this means ppsS is going to be a killer bee.