That is a good point about the RFC. A report filed at
Bug Reporting - Apple Developer is about the only way to try to get the message across. I'd note, though, that things are gradually changing in the CalDAV world to become more "server-centric". While
draft-desruisseaux-caldav-sched-09 - CalDAV Scheduling Extensions to WebDAV doesn't have much to say about them directly, the thrust of the draft RFE is, I believe, to let servers do more of the work. "External" attachments arguably make more sense in this paradigm.