APA doesn't know either. You might think that a scientific field would appreciate parsimony and apply that notion to their stupid citation rules, but you'd be wrong.

Here's a good one: if a journal is sequentially numbered, include just the volume and leave out the issue number. How stupid is that? I've always refused to do that. Why make it harder to find the damn article? And in the 6th edition, they changed their mind, so now everyone's database (except mine!) is wrong. I've also been putting DOIs in my database for a long time. Apparently APA has decided that they're important now, but, really, is everyone supposed to go back and include them now for every article they've ever referenced? Do you have to go and check to see if a journal has gotten on board with the whole DOI thing every time you submit an article?

I've never understood why people are so hung up on it. I've never bought an APA manual. Maybe the fact that the APA can't follow it will teach people that it just doesn't matter. Thankfully apa.cls and apacite do a good enough job to suit me.

On the other hand, I do get annoyed when people (read students) don't come close to getting it right. I don't much care if you know when to et al, but cite stuff in the text with some authors and a year in a place that is least disruptive to the flow of the text, but still where I can tell who's work you're talking about. And include enough information in the citation that I can find the article, OK?