Firstly: back when Mearls joined WotC, someone called that as the moment they started seriously working on 4e. Two years later, 4e hit. To me, this looks like a really strong indication that they're about to start serious work on 5e.
Now, about the column itself: Actually, I disagree that the rules can't help a DM spice up his NPCs. 4e monster design has shown that you can differentiate the flavour of monsters through rules elements - see things like the kobold 'shifty' ability for an example. To do the same for NPCs you'd need a large number of 'traits' each with a game-mechanical effect, but that's no bad thing.
Actually, there's probably a space for an entire not-quite-monster book there, giving the 'chassis' for a bunch of NPC archetypes at various levels (the knight, the wizard, the bandit...), and then a whole bunch of traits for those characters. To create a quick NPC, pick the chassis that seems most appropriate at the level that is appropriate, and then customise with two or three traits...
But then, I also disagree with his list of what makes a good DM (and, indeed, last week's list of what makes a good player). While all of the things he lists are true, there's more to it than that. You can be a good DM while missing one or more of these elements, and hitting all of those bullet points is no guarantee of being a good DM. It just seems... soulless, trying to pin down something that can't be readily defined (but that we all know when we see it).
So... I dunno. I'm more confident of the future of D&D than I have been for some time. I genuinely get the feeling that Mearls really does 'get it', even if I don't necessarily agree with his presentation, and the appointment of Monte is definitely a good step. And certainly, the goal of making the best RPG possible is definitely laudible. I guess we'll see.