In terms of game design:
Make sure you clearly define what the game is about and then make rules that produce play that lines up with that choice. Though if you think about it, separating different goals into three categories might not be that helpful as you can simulate genre with the same techniques you can use to expound on a theme. Or you can meet challenges with the same mechanics as you use to produce plot.
In terms of fixing problematic play:
Find out what everyone wants, talk about it and see if you can't get on the same page. Although this one is more of a big model thing of which GNS is only a unnecessary component. It ends up getting in the way as people pick one of the three letters and identify with it or try to find a quiz to tell them how much of each they really are.
To bring this back on topic for D&D Next, I don't think there's anything we can do with it at all. It's not out yet. And if you use GNS jargon in any playtest or feedback reports you send to WotC, I'm guessing it'll be pretty much ignored.
When the game finally is out (even in playtest) I think it'd probably be more useful to figure out what it actually does at the table and see if you like that than to try ot disect the text and figure out which category we can argue for it to belong in.