If you read that, read the opposing point of view that he links to as well:
http://codecs.multimedia.cx/?p=370
You can basically sum everything up as "one big mess" in my opinion.
There's talented coders on both sides -- too bad they didn't find a
way to keep the project together. Strong coding skills seem to often
have an inverse relationship with strong social skills, unfortunately.
For AVbin 9 and 10 we're going to try Libav. Whether or not we decide
to support FFmpeg as well or instead in the future just depends on how
things work out. It makes my life harder, unfortunately, since
there's not a clear leading project that I can commit to tracking
long-term. Because of the limited scope of AVbin (just
decoding/demuxing using a few API calls) we *might* be able to
maintain support for both.
~ Nathan