For the record, though this post is old, I also have this issue. It is not editors beng whimsical, there are good reasons for it. If you go to 2 authors (Jones & Smith 2007) but use et al for 3 (Jones et al 2009), then it makes sense to sort by author 2 for 2 author papers, because both authors are given in the citation, but my date for 1 or 3+ authors because no second author is given. I agree that the exisiting options get us close, only the custodians of the styles can say if it is close enough.