We are speaking about a very, very simple industrial machine. It cuts the film with a blade - just like the 120 cutter. Or like scissors. The "magic" is to roll it on a core with a backing paper. Yes, it's a bit different from 120 machine because there are two sheets of paper and two pieces of tape instead of one. Such a machine (for 120, which is very similar) is shown in a KODAK documentary film from the 50's. The link has been here on APUG. The machine shown on the film is not big at all. I would call it compact!
If the machine is big, then it produces a vast amount of rolls. If they all sell, then the big machine is justified financially. If less rolls are needed, then there's absolutely no reason to buy such a big machine. This is a self-controlling system. Someone just have to make the correct decision. It's not rocket science but very simple industrial business.
This is very, very simple automation nowadays. In fact, this is about what happened in industrial revolution in the 18th century. The machinery in question is not complicated, and not big even at industrial scale. You cannot even mention it in the same sentence with emulsion manufacturing, coating etc. - those are really complicated and costly operations. 120 vs. 220 is just a matter of packaging, which has been automatized centuries ago and has nothing special, magic or costly here. Machines are used because they are much cheaper than people doing the same job. One person could roll 220 rolls in 10 seconds or so. It wouldn't be so costly even hand-made. Cutting film and inserting a BACKING PAPER is an extremely simple task.
Sorry, but I respect clear logical thinking and facts I can find higher than authorities named but not citated. It's too easy to mis-interpret and read off the context.
I'm 100% sure that the actual problem lies in marketing strategies, like: how many rolls of 120 and 220 should be sold and where (it's difficult to foresee the exact sales number---leftovers are not wanted), or simply a problem of market places (certain stores may have a fixed number of places for products on their shelves and they would be reluctant to take any extra products for sale), etc. Things like these. This is the reason I'm saying that the end-chain needs some optimization.
And, as I said, it needs some reforming anyway! Film business has changed much and it needs continuous care and renewal to stay profitable. One example of such development is the demise of small, local stores, often criticized here at APUG, but it's the only way the FILM can stay, and that's what really counts. As you can see, large Internet/mail-based companies do not have the problem I mentioned about 220 - fixed number of products on their shelves. And, as they sell more, at a larger geographic area, the problem of leftovers due to fluctuations of market is less pronounced. This kind of development might well make 220 more successful. The question is, is it still going to be successful enough to be produced? What else can be done, if anything?