This is a much different definition of Autapomorphy than I am used to, which is a character state that is unique to a terminal taxon and thus uninformative in a phylogentic analysis. I believe this is the definition that David Peter's is also following when it describes them as "weird". I would think that the term apomorphy is what you described above.
I'm using Hennig's definitions, under which one taxon (no matter how big) has autapomorphies and two sister-groups have synapomorphies; those synapomorphies are autapomorphies of the smallest clade that contains the two sister-groups. So, fused frontals are an autapomorphy of Pachypleurosauridae, Carcharodontosauridae, and Anthropoidea _each_.
Hennig was one of those people who like to make up terminology for the fun of it, but this way the meanings of these terms don't depend on the taxon sample of a data matrix. If you want to say a character state is unique, just do that.