I asked this question and the package author replied. Someone suggested that this has to be the right answer. By definition, almost! But there's a bounty on this question, and it seems a shame that lockstep's work at hacking together a working solution goes unrewarded.
Any advice on who should get the bounty?
Luck of the draw.
Whenever I submit a hacky answer to something then I know that I'm running the risk that someone's going to come along and say, "Just use package Y with option X and it all works Just Right.". Then all my hackery is for nothing. Except that it's not, I'll've learnt something by doing the hack, and there's a sizeable chance that there isn't such a package, and even if there is if the person needs an answer Right Now then waiting for Package Right to come along may take too long.
So I make it a personal policy to vote up "hacks" even when a later answer supersedes them. If we get to the point where we're all waiting for an "Expert" to drop by then the site will stagnate and die. But when an expert does drop by with what is without-doubt the right answer, then we commiserate with the unlucky loser and say, "Better luck next time".
I would say that you should only accept the answer after the update is on CTAN! If it doesn't get updated in time, lockstep gets the bounty as that's the most usable solution at the time that the bounty is awarded. And if the package maintainer gets the bounty, that's quite a nice "welcome package" that will encourage him to check the site again from time to time.
My solution was hacky, and I don't think Phil Kime will use it as a base for patching biblatex-apa. Therefore, I concur with Andrew Stacey: If biblatex-apa does get updated in due time, the bounty should be awarded to Phil Kime.
