Perhaps I'm an outlier -- a sole proprietor, freelance designer and photographer who depends on many Adobe products to make a living. To keep up with what my clients need over the past 5 years since I went freelance, I've had to get highly proficient with web, interactive, digital publishing and motion graphics quite in addition to the standard print tools. For me, as for most designers, Photoshop is just one tool out of many -- "a great plug-in for InDesign" is how it's often described. I've also never considered Adobe software expensive, especially compared with professional-grade tools in other fields. One seat of Maya or AutoCAD costs more than two seats of the entire Master Collection, for example. Given their dominance, Adobe could easily charge far more.
I subscribed to the Creative Cloud when it launched, and I've been more than happy. I have a CS6 Master Collection perpetual license, too, but there were advantages to the CC approach and I had already decided to stick with it going forward. Besides the nice, predictable monthly budget, the whole installation, upgrade and management process has been seamless. Anyone with experience installing Adobe software knows THAT is a major advance in its own right.
Notably absent from this thread are two factors that more or less forced Adobe into a subscription model in the first place, and the possibility that there are might be sound human and technical reasons for it. The first is legal, the other two derive from the decision that professional customers would be much better served if all of the products in Adobe's line-up worked together as a suite, rather than as disassociated individual products. We tend not to think about it much, but I have enough years of management experience to know that the logistics and coordination involved in getting more than fifty teams, spread across the US, Europe and assorted bits of Asia, to all finish their release cycles at the same time is mind-boggling. But the suite approach made that essential.
Adobe's wake-up call was the digital publishing revolution, kicked off by the success of the iPad. Customers (me included) were screaming for tools that the Creative Suite didn't have. The accelerating pace of change in video and web technology was already creating a strain; Adobe had been forced to shorten their release cycles from two years to eighteen months to keep their products relevant. Now their digital publishing and web tools needed immediate upgrades also. The obvious answer would have been to push out a feature upgrade, but legal reasons to do with revenue recognition (for which you can thank Enron and others) prohibit adding functionality to an already-shipping product. I don't know all the details of the mad scramble that ensued, but I do remember that it was a scramble, from which we got CS5.5. If you wondered why there was no upgrade to Photoshop or Illustrator in 5.5, now you know: it was all about ePub and Digital Publishing. Subsequently, Adobe announced they would be shifting to a yearly upgrade cycle. From their perspective, it was essential.
Consider the human strain on the engineering teams who have to meet these kinds of deadlines. Arbitrary deadlines have a way of stressing people out. People rushed and under stress are more error-prone. Every suite version has shipped with "known issues" that showed up too late to fix before release, and features the teams wanted to implement but had to defer for lack of time. If you know anything about software engineers, and Adobe engineers in particular, you know that this situation was probably doing nothing to improve morale.
Assume (because by my observation it's true) that from the top down, these people are focused on making the best tools possible for professional creatives. Assume, too, that management necessarily cares about their people, who are plenty talented enough to find jobs elsewhere. Management certainly knows Adobe has to innovate like crazy just to stay in the same place, but they can't burn out their best people with impossible development schedules.
A subscription model solves "How do we keep customers up to date without burning out our most creative engineers?". Personally, I think it was a pretty ballsy decision. It couldn't have been cheap. When they announced Creative Cloud in 2011 no-one -- I mean nobody -- I talked to at Adobe knew if it was going to be a success or a major flop. There were a lot of nervously-optimistic conversations going on at MAX that year. I also remember the delight and astonishment a few months later, when Creative Cloud launched and was an instant hit.
Fast forward a year. Cloud delivery really did work well, updates had been finalized and pushed out via the cloud, more and more customers were signing up, and it became clear they had a hit on their hands. With no suite-wide deadlines to meet, the different engineering teams could schedule their feature builds and QE work to fit the need -- form following function rather than the other way round. Don't think that blood pressures weren't going down all over the place.
Consider the whole DVD manufacturing and packaging evolution that has to happen every time there's a new set of boxed products -- two design suites, two web suites, a video suite, a master collection and every individual point product. All that is now gone, and good riddance I'm sure. Replacing it is the ongoing cost of maintaining all that cloud infrastructure. Like the subscription itself, a one-off big cost is replaced by an ongoing smaller one.
It is a BIG mistake, imo, to assume that a departure as radical as Creative Cloud could be thought all the way through for all possible scenarios in as little as a year. It's going to remain a work in progress for a while yet. I don't accept, because it's contrary to all my own experience, that anyone at Adobe wants to screw over their customers. That's just paranoia, frankly: popular in some quarters, but myopic and misguided. I've seen plenty companies where that was true. The signs are unmistakable: customer service goes down the tubes, all the best people leave, the company tanks frighteningly fast. If anyone remembers Ashton-Tate, that was the perfect example. I don't see any of those signs at Adobe.
I think a subscription model was unavoidable, just as I know there will have to be accommodations for not-very-well-thought-out situations. Photographers who use only Lightroom and some Photoshop, and who feel dumped on by the subscription-only model are a case in point.
Perhaps photographers are a minority public for Photoshop, which is used in every design, film and video studio as well as in forensics, astronomy, medicine and many other fields, but if there's one thing I've learned in the years I've been involved with Adobe products it's that their people do listen, and do genuinely care how useful their products are to all their constituents. The best approach from the customer point of view is to make the case, clearly and without venom, why a subscription model doesn't work. Don't assume, though, that your particular situation is intuitively obvious to an outsider, applies to everyone or is being deliberately ignored.