I actually prefer the "when it's done" model rather than the "train leaving the station" model, for software like this. I feel it better allows for broad-scope changes, which may touch large swaths of a codebase, to more fully gel/mature before being released. For something that I might only update once every few years anyway (ie: not very security-critical), it seems worthwhile to reduce the overhead of release process/logistics for the developers, letting them spend that time on features/fixes instead.
Or maybe there was another model you were suggesting? What do you find non-sane about the current one?
They basically stopped doing the previously-usual updates for a 7 years to focus on porting to GTK+ 3; GIMP was very tied to GTK+ 2. There were a bunch of much-anticipated features (esp. non-destructive editing) that were finished but couldn't be released because it was half-way through the big GTK transition.
Hopefully the now-impending port to GTK 4 will be a lot smoother and won't be such a disruption to their ability to ship the features they've been working on.
People seem to forget that GTK was originally created for GIMP, Gnome came around and co-opted it since it was more free (libre) than QT.
GTK3 was a full rewrite divorced from the GIMP development cycle so broke a huge chunk of things GIMP put in place specifically for their code styling. Thus the long development cycle.
The problem, like most things with GNU itself, is that large dreams are thought of but never actualized. So, just like using the Linux Kernel vs Hurd, the GNOME project had to plugin what was available; and that was GTK.
If those fixes take years to get released, that's not really a good thing. One year with bug fixes is typically okay unless you're dealing with hardware.
From the release notes: "we also intend for minor releases to be much more frequent. Rather than having another 6+ years development schedule for GIMP 3.2, we plan to release it within a year of 3.0".
Hopefully they'll adopt a more sane release strategy going forward.