This paper presents exploratory work investigating the suitability of the Music Ontology - the most widely used formal specification of the music domain - for modelling non-Western musical traditions. Four contrasting case studies from a variety of musical cultures are analysed: Dutch folk song research, reconstructive performance of rural Russian traditions, contemporary performance and composition of Persian classical music, and recreational use of a personal world music collection. We propose semantic models describing the respective do- mains and examine the applications of the Music Ontology for these case studies: which concepts can be successfully reused, where they need adjustments, and which parts of the reality in these case studies are not covered by the Mu- sic Ontology. The variety of traditions, contexts and modelling goals covered by our case studies sheds light on the generality of the Music Ontology and on the limits of generalisation "for all musics" that could be aspired for on the Semantic Web.