Adam Lindsay
It means that there's no \omacron in texnansi (or the other major western encodings used for TeX fonts), and you're asking ConTeXt to synthesise it from an 'o' and a 'textmacron'. It wouldn't surprise me one bit to hear that the Mac handles combining accents differently from TeX, and so the converted font's metrics is going to result in mistakes in building text accents.
Interpretation: you're using artificial test cases to verify the font installation. Stick to what you need and know, and you'll probably end up with a character in the encoding.
No! I do need \omacron as I am typesetting the German translation of a Indian Yoga book ... with lots of Sanskrit (in latin alphabet plus all those diacritical accents)!
If you do need \omacron and the like, then you could hand-tune macros that accommodate the specific font--see Idris's recent thread on accents, or you could cook up an "expert" encoding (since MacOSX's Helvetica does include that character) like Vit, and switch to that encoding when you need the character, or you could even go with a big unicode font install. I'll be documenting that process sometime.
That would be quite interesting. (But for now I leave this workshop and use the HelveticaNeue to go on.) Thank you, Steffen