Freeze synchronisation rather than Release synchronisation

I must say I'm a tad disappointed by the coverage of this one.
Normally LWN gets it pretty much right but here something very important was missed.

This is not about syncing releases. It's about syncing on freeze points so that enterprisey
distro releases start off with rougly the same base versions of software in order to increase
the test coverage and drive down costs at the same time. For instance, if upstream knows the
major distro releases are about to pick one of the latest releases as "the" stable version to
ship they can plan to get something with a bit higher QA level out there for them to pick up.

So if I understand this plan correctly the release dates could still differ wildly but the
point where the distros are freezing on a set of versions of packages is synced. So you'd get
several distro's releasing at different times with almost identical version levels of packaged
software that would still carry different levels of QA and patching.

And THAT sounds like a very good idea to me ;-)

(Search for "Mark Shuttleworth says" in that article if you want to verify what I say here.
The comments are quite informative.)