I invested quite a lot in Seam 2, and I was (and am) happy to see that the core of it became a standard with CDI (that I really like) but obviously after you have many projects and clients you cannot just "upgrade", so the idea was to keep Seam 2 while the development of Seam 3 was going on, and in the meantime move to JSF2 with the help of Seam 2.3. After that, move everything to Seam 3.
As everyones here know things went differently so Seam 2.3 release was VERY slow (it is not even final now) and in general all the development was moved to Weld and Seam 3. So after a while I've decided to start again with Seam3/JSF2 (and keep the old clients on Seam 2 until the new platform would have worked correclty).
Now after investing in it and putting a lot of time, the official answer is to wait for DeltaSpike and use it when it will be ready...
I have learned a lot in this period and I've fixed a lot of Seam 3 bugs with the help of other persons and by myself.
I know that's the risk of open source etc. but I think you can't just say again "wait"...DeltaSpike will not be productor ready for another year and I (and I'm sure many like me) simply can't wait.
I also understand that RedHat/JBoss is not committed on this stuff like it was one time and that's fine because at the end of the day you have to follow the business decision but AT LEAST accept the fixes from the community and do some minor releases that address these fixes so that Seam 3 would work properly while everyone is waiting for DeltaSpike, I don't think it would be a so big effort and anyway that's better then nothing (I guess everyone working with Seam 3 has developed his own fixes and workaround).
In the Seam website you say that Coding has become social (http://www.seamframework.org/Seam3/Contribute) but on the link above (https://github.com/seam/faces/pull/75#issuecomment-8467066) what I can see is a guy with a possible fix (that worked for me as well, even if I will do some changes to it) and no answer about reviewing and/or integrating that fix.
I think that Seam 3 need at least other 2 minor version updates to address at least the problems fixed by the community, also because I think it's RedHat's interests to keep the community around it instead of having it moving to other projects (it would be a waste for all the good things and community Seam has brought so far)...
I really would like to hear other opinions about this topic...