*Ceratops*: has anyone even looked at the material in the last couple of decades?
Taxonomy is not just about compiling lists. It's a way of classifying real species. If we cannot demonstrate that _S. armatus_ was a separate species (because the type material cannot be distinguished from diagnostic _Stegosaurus_ species) then the taxon is toast. P here's no good evidence that this ever represented a real species, then let it go.
"Real", or "species"?You imply a morphological species concept here (or, more likely, all of them). What if I happen to prefer a "biological" or ecological one? Then I'm out of luck, because such concepts cannot be applied to stegosaurids as we know them today (and indeed to almost all extinct organisms). In that case, I can either pretend that some morphological concept or other is a good approximation of whatever I prefer, but then I'd be lying to myself*; or I can stop worrying about species and talk about morphological LITUs** instead. _Except_ that the ICZN doesn't let me choose the second option. It stupidly insists that every organism that is to be classified at all must be referred to a species. The PhyloCode can't come soon enough.
* Depending on the species concept, there are from 101 to 249 endemic bird species in Mexico. A factor of almost 2.5.
** Least Inclusive Taxonomic Units -- the smallest recognizable clades.