I imagine BDFL would do anything at this point to get Py3k jumpstarted.
This should give it a helpful kick in the pants.
Rails is just a web framework. Killer? Yes. All you need for many apps? No.
And yes, both Django and numpy are also killer packages ... with Django probably being the inspiration for merging Rails with Merb.
Much of the joy in working with python comes from the "already written" effect. Libraries exist for almost any imaginable task, and it doesn't cease to surprise me how good the average code-quality is. It's extremely rare for me to run into a foul egg, which is a pleasant change from the platforms I've previously worked with.
My point is only that python has no similar library or framework driving uptake.
Only because the python community is not as vocal as, say, the rails community doesn't mean it has no uptake.
Nothing happens overnight, this is just something to incentivise possible porting efforts.
That said, I don't see this as a tactic to get people to upgrade. Python is an open source project and the developers get to choose what they will work on. If the Unladen Swallow guys want to merge into 3.x, that's their right. If someone else steps up to merge it into 2.x and support it, that's their right too.
Myself, I'm going to continue to fix bugs in the 2.x stream for the foreseeable future. I do plan on moving my applications to 3.x eventually though.
Unladen would require a whole new 2.x release, which is not, and more than likely will not be planned. We're not playing games, we're putting features into 3k knowing full well we plan on putting 2.x out to pasture. We're making the selling points of Py3k porting much more compelling to library/framework/etc authors.
If someone wants to fork 2.7 and merge it with unladen, or better yet simply run the 2.x release of unladen-swallow up on google code, that's totally their prerogative, and I know some people will. From a feature standpoint though, this is a good think for 3k as it sits today.