I did look at it a few years ago (perhaps the title of your post should be "Is anyone using SKOS"?). For tOKo, I was looking for a standard that includes the definition of terminology (synonyms, abbreviations, etc.), the standard semantic is-a relationship and the notion of narrower for topics. At the time I looked at SKOS, it appeared contrived and lacked any proper interpretation. My conclusion, with similar efforts in mind, e.g. RSS, Dublin Core, FOAF, was that there had to be some critical mass to make it worthwhile to commit to it. And in the end I decided to define my own terminological definitions on top of RDF(S).

For some reason or another to make these kinds of standards work there has to be some critical mass. RSS had this in the beginning, but nowadays if you adhere to the RSS standard your software won't work as the RSS data comes in many, many flavours. The requirements change, for example the introduction of tags, and in the underlying formalism (RDF) it is to easy for anyone to come up with an extension that someone will find useful and consequently can break existing software.

For SKOS, a possible alternative is the WordNet "standard" (http://www.w3.org/TR/wordnet-rdf/). At least WordNet comes with a large dataset, which makes it more attractive to commit to. I have not followed SKOS developments lately, but doubt there is a sizable community of followers.

Anjo.