There's a pretty large amount of 0.46 bugs being reported again and again in launchpad, and most of them are fixed in SVN. I guess that stabilizing the SVN and fixing important bugs and regressions to release Inkscape 0.47 as soon as possible would have a significant impact (positive) in the reduction of duplicated bugs.

I personally don't care if some filters, lpes or features don't make it. I'd prefer a stable 0.47 with the bugfixes and the enhancements that may be considered "stable". Even with its regressions and some high importance bugs, the SVN version is pretty stable and it has loads of improvements (snapping, pdf output, eraser, spiro, etc.) that would make 0.47 attractive enough, from the user pov. And that would help to close lots of fixed bugs for good.

So +1 for the idea of starting the freezing and regressions hunt.

Gez