Interesting. I do not have a check valve between the reactor and my inline bubble counter.
When the water in the bubble counter gets low, I loosen the top of the bubble counter and water flows from the reactor thru the cO2 tubing back to the bubble counter and fills it up. When i turn the cO2 back on, I can see the gas push the water in the tubing back into the reactor. Its not super fast movement, but in only takes 30 seconds or so to go 2'. Again, I am running 20psi.
Given how you installed your tubing into the reactor, the reactor "may" have a greater influence on the operating pressure of the cO2. In my system (early part of video), you can see my cO2 is injected perpendicular to the flow of the water. Does this lower my need for psi?? Don't know. I would think psi in the tubing and reactor are the same and therefor irrelevant between our systems.
"I can even see the level bobbing a little bit with the flow of the water from the filter."
Interesting, i would think at 20-30psi the cO2 would simply push the water right back into the reactor. This statement does seem to make me believe your cO2 is "not" going into the reactor.
"When I decrease the flow, the bubbles stop moving," Decreasing the "flow" or bps should only effect how fast the bubbles get to the reactor (more cO2 volume within a given space - the tubing). Your psi is unchanged so I would assume the rate of cO2 movement would slow down, but is should not stop.
if you fill the tubing with tank water, set the regulator at 20 psi, then adjust your pbs to something higher like almost a stead stream of bubbles, what happens to the water in the tubing?
This is my flow rate