[tantek]aaronpk, you're right about this part: "spec distinction is useful to allow each part to develop independently", specifically *when* that "useful[ness]" is defined by user-feature usefulness, not spec-author usefulness (see also priority of constituencies)
