I always wondered if the ω glyph developed when lazy monks sloppily doubled up oo to make the o-mega… but I never investigated whether this is true.
Also, from my understanding, ω is just a degenerate form of the upper-case form (Ω) when drawn really flat and with the edges flaired up [1].
Ω is the capital of ω, exactly the same letters, though ancient Greeks only used capital letters.
The Ω was pronounced a bit more deep in the throat, kind of like "ho", but that difference was more accented when the letter was used by itself as an article "Ω human", for example.
Still though there is no absolute knowledge for the exact pronunciation of any of the ancient Greek. My belief is that the most close modern dialect is the Cypriot since it's the most melodic and rhythmic.
They had different letters for the same sound but different duration, such as Omega (ω) an Omikron (o) which means O-big and O-small.
The only problem is that we don't really know the exact vocalization of all the various letters and notes (notes as combinations of letters, for example ΟΙ or ΕΙ) which have the same vocalization in new Greek. We do have however some clues as to how they were pronounced.
All in all, because I have been following the subject for some time, we can't reproduce the music as it was. We have some directions as to how to rebuild the instruments and that's the best lead up to date, but still that's up for much debate as to how accurate the reconstruction is.