Yep, you are correct that a pastel (i.e.) is het for that mutation. It just becomes difficult when someone says "het for pastel" because that implies that two mutated alleles make a pastel. But, a pastel is a het (het for that chromatophore anomaly).
And, yes, in my career I've always done the cap/small letter (i.e., Pp) to indicate one mutated allele for (let's say) pastel and the other unmutated homologous allele (homologous roughly meaning paired or same location). So, for those who need an example, PpHh would be a pastel het for ghost (hypo), and PpHH would be a pastel ghost--you can make up your own letters. It could be helpful for some to insert "n" where there is a normal allele, but for me, it's just too busy. So Randy, Corvid, Dr. Del....good posts. I concur.