John,
In my tests with the setup shown, I found even development without side-to-side agitation. I cut open-ended film tubes from 1.5 inch ABS, 15.5 cm long. I chose that length so that the 4 inch dimension of the film, when centered in the tube, appeared to be sufficiently far from the turbulence at the end of the tube (which would cause increased development), so that the entire sheet gets something closer to laminar flow. I cut a 4 inch PVC pipe section to 20.5 cm long and glued end caps with centered 5.6 cm diameter holes over both ends, large enough to easily insert a 1.5 inch pipe and extract it easily in the dark. These are not true 4 inch end caps that fit over the outside of the 4 inch pipe, but might be called test caps or something similar, and fit inside the pipe. Be sure to glue the end faces of the 4 inch PVC to the end cap securely, not just the interior of the pipe. The 4 inch pipe interior has small facets and won't seal completely. I run this setup on a Unicolor roller base, and can use reversing or single direction rotation. My setup takes something like 150 ml to cover the film, and the small tube rolls inside the large tube.
I also cut some 18 cm lengths of 2 inch ABS pipe and put similar end caps (with no holes cut) on one end of those. I can float one of the 1.5 inch ABS tubes inside a 2 inch tube and have 250 ml cover a 4x5 sheet. It floats and bobs, and the 1.5 inch pipe floats higher than the top edge of the 2 inch pipe, so you can just tap and spin the 1.5 inch film tube to agitate. I do this with the film near the bottom of the small tube so that it's completely submerged. I'm just starting to test this with semi-stand agitation, but see no reason that it shouldn't work.
Hope this helps. My main point is that I think you can get even development without having the tube tilt end-to-end. You might try it before making something more complex.
Lee