Some random thoughts...
It's only my opinion, but I think Java 1.5 and SEAM are important for the project. From a marketing perspective, it says, "We're on board and part of a unified technological front--Jboss is/is going to be the leading JEE5 solution provider, and we're a major part of all the inovation that brings."
If a company, large or small, is going to adopt jBPM, they're doing it because they're ready to do something new--and a Java 1.5 requirement isn't likely to stop them.
Working for a big company, I know that we'll choose a solution based on its own merits, regardless of which JVM it runs on (within reason). If we choose the latest jBPM, we're running it on its own system, and on that system we'll run whatever is required. --we just enjoy working with 1.5 better, so we do :)
Looking at web technologies, its been our experience that JSF, on its own, is a royal pain--so we're moving our task management web app to Facelets (and perhaps SEAM) looking for better productivity. If we can't find it, we're likely to have to revert back to straight JSPs or be forced to a non-Java technology. That is just to say, that it's hard to go wrong by bringing in strategies that increase productivity and maintainability.
Also, when you find a group of programmers who are excited about a technology (like Gavin is about SEAM) it can be a lot easier to find someone willing to help work through sticky problems.
If we go with Seam before 3.2.0, I'm sure it will delay the release of that version. That, and losing jdk1.4 compatibility, are really the only negatives that I see.
As far as technical merits, JDK5 is obviously superior in lots of ways, and I think going to Seam would be a smart move on many levels.
Let's do it?
no not yet. let's first wait and see if it actually brings the extra contributions we're hoping for.
i think that for 3.2. we need to go ahead and finish with what we have now.
I agree, lets keep 3.2 as it is. 3.3 maybe... Maybe we should do a poll in the userforum?