At that time, Java3D was the fastest, jME came behind it and Xith was the slowest by far (so much for tuned Java3D...).
However, all engines have evolved since then, including jPCT. So this test may be no longer valid...
Correct, it is no longer valid. I believe I know which comparison you're talking about. It's over at the jME forum. That discussion has greatly evolved since you last checked. It's now clear that jME is the fastest engine:
http://www.jmonkeyengine.com/jmeforum/index.php?topic=2597.75jPCT does a lot of things in software while the other engines can do this hardware
So would it be safe to say that the software processing of jPCT hinders its performance? I do understand that this is a trade-off for a more dev-friendly API but nevertheless it still sounds like a sort of performance bottleneck.
You can add and remove stages from single polygons at runtime. If you want to add 20 texture layers to a single polygon if something has hit the polygon, you can easily do it. You can't do that in any other engine.
From what I'm understanding here, jPCT's biggest strengths are in the texture management system? Please correct me if I'm wrong, I'd really like to understand this.
you can use the normal GLRenderer as well as the AWTGLRenderer in the same way
Yes, this is definitely a BIG PLUS!
You don't have to care about threads or put every jPCT related command into a gameloop-construction like you've to do in jME
I'm not so sure about this, I've had some glitchy behavior if any objects in my scene were handled (i.e. manipulated) outside the main jPCT game loop.