The plans we had grew bigger, and more complicated… what we have might not be exactly what people expect, but I think it’s just going to be seen as objectively better.
So what does that mean?
No idea, honestly. I think, however, we can safely rule out a blanket open license a la 3e’s OGL.
One thing I picked up from an interview with Ryan Dancey (I think on the Fear the Boot podcast) was that the OGL didn’t quite work the way they expected. Dancey and Co. had assumed that DMs would use the OGL to publish their homebrew adventures. (Keep in mind that, at this time, adventures were seen as loss-leaders; necessary support to grow an RPG, but individually unprofitable.) Instead, what they got was a flood of splatbooks.
And a flood of new character options is absolutely not what WotC wants to see for 5e. If you got back to Mearls’ comments from GAMA or the early part of the Tome Show interview, you’ll hear him talk about how important it is to keep the game lean, to not drown players (and especially DMs) with lots of new options, special cases, and new mechanics.
So I’m predicting a license that discourages character options, alternate games (i.e. True20 or Fading Suns d20), and the like, but encourages publishing adventures, settings, backgrounds, monsters and treasure. The push will be to use backgrounds, and not new classes or class paths, to make characters better fit the setting. Maybe races. Races could go either way, but I suspect they’ll be discouraged as well. And spells? Probably allowed, but that’s a grey area that might fall too close to making new classes and class variants.