So the issue seems to be whether the solution volumes listed by Jobo are sufficient both to physically submerge the film loaded on its reels within the dev tanks, and to chemically process that amount of emulsion area. The solution bottles hold 750mL, but no process run by the Jobo requires more than 660mL of solution. Well, I located and dragged out the manual for my ATL-1000/1500 (they are identical except for the control head) to try to bring some (further) clarity to this thread. I don't have a digital version, nor the capability at present to scan and post it. So bear with me through a tedious post.
Section 4.3, "Determining Chemical Amount in Solution Bottles", talks about two ways to determine required solution volume:
a more precise method of measured amounts of chemistry, by reference to a table on page 24 in the manual, based on the desired number of rolls or sheets. Looking at your beloved 220 rolls, the table indicates that to process three rolls of 220 (the most you can do in one batch) requires the larger tank and 660mL of measured solution, while to do three 120 rolls requires the larger tank and 330mL solution. This works out to about 110mL per 120 roll "equivalent". So this is what Jobo says it will take to submerge and to process a 120 roll equivalent (two 120 = one 220.)
a quicker method using one of three standardized solution volumes: 170mL, 300mL, or 640mL, determined by looking at a sight glass in each chem bottle that darkens as it is filled. Half dark = 170mL, all dark = 300mL, filled to 1 inch below the bottle neck = 640mL. These three amounts will cover every permutation of film and tank size. The quick method is what is referenced by the illustration printed on a sticker atop the machine itself. This sticker lists fewer permutations of tank and film as befits its simpler orientation. The quick method may require slightly more solution for a given roll/tank combo than the measurement method in certain instances, but is easier and quicker to use.
I've always used the quick method, and it's never failed me with either C-41 or E-6. The simplified sticker table lists nothing for for 220 film, but does list something for six 120's, which when loaded on the three reels is exactly the same thing (two 120 per reel = one 220 per reel, identical.) For this, you need the maximum-full setting of 640mL, while for three rolls of 120 you need the large tank and 300mL solution (sight glass all dark, ie, intermediate fill capacity.) So by either method of filling, Jobo says you need 100-110 mL solution to physically cover, and to chemically process, a 120 roll equivalent. Pretty close.
The question seems to be, then, whether this 100-110 mL of solution is enough solution to actually process the film, even if it's enough to submerge/cover the film.
Enter Kodak Publication J-83 and TI2443 for E-6, and Z-131 for C-41. For E-6, there are different solution capacities listed between the two publications, and depending on which kit (5L or 3.8L) you use, so it gets quite confusing. Depending on which set of numbers you use, they work out to between 125-250mL solution per roll to process the film (issues of submerging it aside) without time compensation, ie, using a first dev time 6:00 in a rotary processor.
You can stretch this by increasing the first dev time to 6:30, such that you effectively require less solution per roll. Interestingly, the Jobo's default E-6 first dev time is 6:30, so it's preconfigured to stretch the solution capacity via time compensation. That calculation works out to about 115-170mL per roll, depending on which set of numbers you use. Given that there's more than a 100% variance in the times listed, there must be a lot of wiggle room built into E-6 solution capacities. In any case, it's evident that Jobo has pushed the efficiency of chemistry use to its absolute maximum. But at least for E-6, there's sufficient overlap between Kodak's and Jobo's numbers to accept that it all works.
The situation is somewhat wierder for C-41. Z-131 seems to indicate that a liter will do only up to about three 120 rolls, which implies around 330 mL per roll. This is significantly more solution to chemically develop than the Jobo calls for, yet it all seems to work beautifully. I regularly process a 120 roll with less than 150mL of solution, with perfect results. I've come across web discussion threads calling into question the accuracy of Kodak's capacity information in z-131; evidently stated capacities have decreased since earlier versions of the information. Further, the Tetenal "press" kits for C-41 require only about 120mL per roll, roughly three times the capacity of the Kodak chems. So not entirely clear what's afoot here.
Bottom line, despite all the uncertainties, processing either C-41 or E-6 in a Jobo ATL-1000/1500 yields flawless results using the chemical volumes Jobo calls for. It seems safe to say that those are enough, but that 220 rolls require about twice the chemistry to do the job. I don't have any further explanation to offer.
: 
