It seems like we're getting way off the topic, but it's fun reading I guess.
I'm not really sure why there can't be a middle ground on all these issues.
Anyway, I stopped by a local shop to get some price ranges. I asked for some very rough figures and knew they were going to be quite high.
If someone made up a CAD drawing, it would take about $900-$1200 to get a prototype made using 3D printing or milling a block of softer material like wood or plastic. If the prototype was tested to be good, and we planned to make under 500 parts or so, then it would make sense to make the cheaper type molds which would run about $500. But I could do that part for a bit less on my own. I'm guessing after that, each one could be made for around $25, depending on the amount of plastic used. The larger the WG, the more material of course.
Now if we figured more than 1000 could be made up, then he recommended injection molding of course. I know that's expensive because I've looked into those before. Rough estimate on the aluminum molds for the machine would be around $15,000. That of course would be after a prototype was made and tested.
The cheaper mold is fine for less parts, but the material to make each part is fairly expensive. On the other hand, the injection molding material is very cheap, but the mold is very expensive. So in the end, the costs come very close if you're only going to make 1000. Obviously as you went over 1000, the price starts coming down as you can average the tooling fees out over more parts. At that point, the injection molding is better.
IF a really good design could be figured out, then it's still something worth looking into.
I'm going to talk to a friend of a friend tonight that runs a mill machine. He might be able to do me a favor. But obviously this is all AFTER a design is figured out.
I realize my thoughts here don't hold a heck of a lot of water for the design portion, but I'll put them out there anyway. I think the size of the waveguide should stay around 15"-18" or less. If someone wants to go bigger, maybe they can do some massive round overs? No doubt some want to go way bigger. That could be better, I won't argue that. But if very few people want something that size, then going through the entire molding process is almost a waste of time simply because it's never going to sell strong enough to pay for the tooling fees.
This project, as with all mass produced items would have to hit a middle ground so that more people could use it. A 30" waveguide using a 2.5" CD may be incredible, but if no one will buy it, then I don't think it can be pursued for this project.