Maybe a good start would be to define an extensible schema and to collect ideas for that schema (such as Mike Dunn's above). This could be a very useful thing, but publishing a schema will make it acceptable for many uses, so long as it is capable of handling many possibilities, and even allows for extension for the possibilities that are too rare to build-in, or ones that you or the community can't think of. Also, and this *could* go along with the schema, using a namespace with namespace-qualified elements (perhaps even qualified attributes, like RDF requires), makes XML parsing easier since your elements would be qualified and couldn't get confused with other elements (by the XML parser that understands namespaces in most cases). I think this is a really interesting idea. Good luck!
-----BEGIN GEEK CODE BLOCK----- Version: 3.21 GCS/G/MU d- s: a- C++++ UL@ P++(+++) L+(--) E--- W+++ N++ o+ K? w++++ O- M(+) V? PS-- PE Y++ PGP++ t++@ 5 X+++ R+@ tv+ b(-)>b++ DI++++ D+ G e++>+++ h---* r+++ y+++ -----END GEEK CODE BLOCK-----