The current standard uses a rigid data model. The problem is how to make something more flexible, but still allows widespread easy use. This is a difficult balance with global standards.
Other work in this area includes the Learning Resource Metadata Initiative (LRMI), based on Dublin Core, which come from the library community.
ps: The meeting this time is at a slightly better time, 5:30 am, rather than 4:30 am in Canberra, but it is still early. There are 34 participants (up from 24 participants at the inaugural meeting). Most are from the USA, but a few of us are from around the world. The meeting is using WebEX again. I have been able to get the audio to work reliably and had to dial in for sound, with computer for video. As with other video conferencing systems, the computer client provides a code to use when dial in for audio, to link the voice call with your online identity.