1. I assume I can render the image inside a JPanel (or similar) to allow JButtons, etc next to the scene? Lightweight or Heavyweight?
As you've already mentioned, this is possible with the software renderer. jPCT will render into an Image, so whatever can be done with an Image can be done with the rendered output (getOutputBuffer() in FrameBuffer will help you here). You may try to use OpenGL instead of software and get the rendered output in the same way, but that's most likely very slow and you'll still have an OpenGL-window lying around on your desktop...so that sounds like a bad idea.
2. How do I render a simple 5-sided polygon with a hole somewhere where the polygon, sides and hole are selecteble? Maybe publishing of some of the sample primitives will answer this question?
You have to model the 5-sided polygon with the hole by using triangles. Either by finding an algorithm that can do it or by using some 3d editor. Picking the polygons is easy, picking the sides...no idea at the moment but should be achievable somehow. Picking the hole is currently impossible, because you can't pick "nothing". A workaround would be to build the hole out of transparent polygons and pick that object instead.
3. Since I probably need to draw selectable 3D Lines and Points, I imagine I can use the Sphere primitive for the points but what about the Lines? Maybe use a very thin Cylinder?
You can easily build some primitives for that yourself. A cylinder should work too, but may be polygonal overkill.
4. I will most likely have 3 simultaneous views on the same model. Does jPCT support multiple simultaneous cameras?
Yes, you can do that. Either by modifying the camera itself for each view or by creating yourself 3 cameras and replace the camera in World with the right one for each view. PrimordialSoup did that. You can see the results in this thread:
http://www.jpct.net/forum/viewtopic.php?t=15Hope this helps.