That is only relevant if there is no significant professional Python 3.x development taking place. There is, in fact, significant 3.x development taking place,

That is only relevant if there is no significant professional Python 3.x development taking place. There is, in fact, significant 3.x development taking place, and that number is growing - exactly as one would expect. And by significant, I mean upwards of 40%. So, except for corporate inertia, or the rare can't-live-without libraries that have yet to be ported to 3.x (see: http://py3readiness.org http://py3readiness.org), there are no compelling reasons to stick with Python 2.x for new projects.