The field of fluid dynamics is quite interesting and was my favorite class in college. Flow through a pipe is actually quite difficult to model. Near the walls of the pipe the velocity of the water is almost zero while the velocity in the center is quite fast. What adds to the confusion is that near the walls, where flow is slow, laminar flow prevails. Out in the middle of the tube/pipe/river/ditch the flow is turbulent, which helps to churn up CO2 bubbles, but which also causes more energy loss from friction.
Since the cross-sectional area of the reactor is actually enormous compared to the 5/8" or 1/2" tubing, the velocity through the reactor itself is small, allowing the CO2 bubbles to rise up against the flow until they are quite miniscule. This low velocity also means that resistance to flow is very, very small in the reactor portion of the assembly. In English - you'll get more resistance from 4 inches of tubing than you will from 24 inches of reactor. Nobody worries about an extra 4 inches of tubing. Much more important than the length of the reactor is the transition between tubing and the large pipe. If you have a long, straight, gradual transition there will be MUCH less turbulence & resistance than if you have a short, abrupt, transition with a 90 degree turn in it.
That said, the length isn't all that important once you get to a certain point. Once the bubbles get small enough, they'll move right along with the flow no matter how long the pipe is. Tilting the whole thing a few degrees allows the bubbles to slide out of the high velocity center & climb up along the wall where the counter-flow is slower. This prolongs exposure time and improves the overall efficiency.
It's kinda fun when hobby & science collide - makes you feel like Einstein when something actually works.