WELCOME BACK Steve!
I particularly like the idea of assigning an object identifier for
each body-part, and the idea for using terminaolgy that melds with
X.400 usage. Since RFC-XXXX is inventing new terminology, we should
avoid being different just for the sake of being different.
So, do we need to set up an OID registry for all this, where we would
register any new objects (OID + Definition), plus cite all previously
registered objects that we use, with a pointer to the registered
definition's authoritative text.
This is the sort of thing that SNMP did for registration of irts SMI
and various MIBs. This looks like a close parallel.