This picture is something my students built very quickly from available materials. https://drive.google.com/file/d/1DO0t_s ... sp=sharing
The 1/2" tubing fits a small steel sphere very nicely, and the jump angle can easily be adjusted. I'm worried that the clips they're using would be problematic, because the ball would be obscured by it for a millisecond (possibly violating rule 3.c.). However, that's just from one vantage point; if you move to look down the path of the tube, you'd be able to see the sphere. I guess the rule isn't as unambiguous as I had thought earlier!
I'm wondering if I should let them run with this idea and build a working device using this track attachment strategy, or if we should stop and find a better way to attach the tubes to a frame.