Normally I write extremely long posts, but I will try to be [more] brief with this one (EDIT: well I failed at making it short =/).
-------
The developer's concerns are valid. The pay-once model creates a less predictable revenue stream, making it harder for the developer to make hiring decisions.
It is also harder to balance knowing how much time to spend on creating new features for the existing major version, bug-fixes, and updates, versus how much time to spend on a new major release. If your internal projected major release date is off from your anticipated revenue, then it could dramatically harm the company's chance of success.
Shorterning the major release schedule to one year (as with The Foundry's Modo or Adobe Creative Suite), doesn't solve the problem, but instead magnifies the issue where the developer has to work on "showy" features, deprioritizing ones that make the core more functional and efficient.
This all generates anxiety for the developer, who wants to focus on making a clean, solid product.
-------
However, the customer's concerns are also valid. All the points the community has been making about ownership, even if a false sense of it, are important and should not be ignored. This plays very much into customer psychology.
There is no coincidence that 99 out of the top 100 daily grossing iPhone games are all free to play (the 1 premium one being Minecraft). This paradox works because they employ monetization models that very much understand the customer.
I believe the developer knew very well that this subscription/rental model was going to be unpopular, and I commend them for being up front with it by making it effectively the first thing they announced this year, as it would quickly become the elephant in the room.
To me, it sounds like the developer knows customers don't prefer it, but are hoping that it won't make enough of a difference to dramatically hurt the revenue. It also sounds like they are hopeful they can make a change to bring the plane back up if it starts to dive noticeably.
-------
I personally do not believe it is necessary to "wait and see" when we very well understand the problem and customer psychology, and can find a solution that is better for both customers and the developer.
In other words, perpetual licenses "protect" the customer and rental/subscription models "protect" the developer. There was a great GDC talk about the idea of protecting the customer, revenue, and developer here: http://gdcvault.com/play/1024183/DON-T-CHANGE-A-THING . There are measures that can be taken that protect both the customer and the developer, which in turn protects the revenue.
To do this, the company should figure out the sources of customer anxiety, and then capitalize on methods that relieve this customer anxiety. Here are a couple examples:
1. Customers hate paying shipping (or "postage & handling" as it used to be called). They feel like this is a sneaky hidden price that jumps out at you after you are already hooked.
To solve this, companies just make the price even more hidden by tucking it into the price. We know by research that a customer is willing to spend an even greater total amount if they see "free shipping" than a price + shipping. E.g. a customer is more like to pay $14.99 free shipping than $8.99 + $5 shipping.
2. Many who buy plane tickets are afraid that they may have to cancel their flight, and thus lose lots of money. Customers absolutely hate the feeling of wasting money.
Companies can then exploit that anxiety by figuring out how often customers actually do cancel flights, and then create a flight insurance option, charging well over that amount multiplied by the standard ticket price.
The whole insurance industry is based on customers paying for peace of mind, paying away their anxiety.
-------
Scirra's customers aren't bluffing when they imply they are willing to spend more than they would otherwise in order to eliminate their customer anxiety caused by rental/subscription pricing schemes.
No matter if you do annually, monthly, daily, or hourly payment intervals, it inspires a sense of pressure for subscribers to use the software, or else they are "wasting" money. I don't feel I need to elaborate on all the concerns the community has raised. This thread and others are rife with it.
The solution:
There are all sorts of interesting freemium ideas that could do very well to build the customer base, boost PR, and give the "whales" an opportunity to spend crazy amounts of money on Construct. Unfortunately, there isn't much time to revamp Scirra's systems to support some of the more effective ideas.
Due to time constraints, among other reasons, I believe the best solution right now is to keep the $99/year subscription/rental system already in place, and add a boring 'ole perpetual option/rent-to-own system.
Customers are different, and also myopic when it comes to finances and statistics. Some customers will only buy movies, paying 5 times more than the rental price, even if they watch them each an average of 1.5 times, and never resell them. Some customers pay for expensive cell phone plans which give them "free" upgrades on phone hardware every 6 months (way faster than they would otherwise), so long as they return their phone each time; and some would only buy phones (with either contract rent-to-own or bought outright), feeling they now "own" their phone they'll usually never use again or sell. Either way, the company can exploit them while making them feel satisfied.
I digress. Scirra can predict generally how many years they expect or even hope customers to subscribe each, and when they plan to release the next major version (if they plan to do so). They can then set the rent-to-own timeline to be further out from their hopeful customer retention, and even potentially in line with a distant, projected next major release.
For example, say you can foresee technology and development trends changing enough by 2022 (5 years from now), that you could see it as an opportunity to deploy Construct 4 by around then, employing the development overhaul necessary to meet the new trends.
The maximum Scirra would reasonably get from a single customer would be $500 from Construct 3 before Construct 4. Say Scirra then did a $500 pay outright, and five-year rent-to-own feature (where, any customer who subscribes at $99 per year for 5 years gets to own it perpetually).
If there was an option to continue subscribing for C3 or subscribe to C4, then it would be naive to think any significant number of customers would continue subscribing for C3 when C4 was out. I'd imagine Scirra would likely allow C4 subscribers to use C3 "for free".
All that being the case, the only way Scirra would miss out on a customer (and thus any revenue) would be if there are any customers who are so utterly shocked by seeing a $500 sticker price (which itself could be less advertised) that they would turn away, despite being willing to pay $99/year.
Any other customer of any kind would be at worst, be even with their currently proposedd model (in all practically significant scenarios).
Any customer buying outright would be a net gain in revenue if they bought in 2018 or later, for all practical predictions.
Any customer continuing to subscribe in part due to the rent-to-own dangling carrot would only positively affect net revenue even if they did make it to the end and "owned" C3.
Protect the developer: the revenue would be more consistent than C2 and most definitely greater than with a no-compromise subscription system. The developer can focus on making the engine greater, by their vision, without stressing as much about gimmicky bells and whistle features each year, or stressing about maintaining bug fixes across different minor versions (as with the "keep your most recently paid version" model).
Protect the customer: they aren't "wasting" their money by letting months go by without using the software, since they are "working toward" owning it. Open a revenue stream for those who want to "collect" tools, or the many others who are psychologically deterred by losing their sense of "ownership" in the many different ways it manifests.
--------
My prediction
I'm guessing that Scirra felt that they couldn't have increased the price from $169 (or whatever it was) to $500 between C2 and C3 without an even larger uproar and potential loss of revenue than migrating to a subscription/rental system. They would have only been able to increase the price to something like $300. I'm guessing they felt this target revenue was something they could achieve more easily per customer with a subscription service, despite any outcry.
My prediction is that they are going to "watch and see" what happens this first year, and what happens when they raise the price from $50 to $99 for the C2 owners, to get an idea of the market and how customers respond to the rental/sub model.
I also predict that a different company or two will see this as an opportunity to poach a lot of Scirra's customer base, as Serif has been doing suddenly and very successfully with their Affinity Photo and Designer product line against the 'untouchable' Adobe's customers (note the "No subscription. Just $49.99" emphasis on the front page of Affinity's products).
That pressure, along with [what I anticipate to be] a disappointing subscriber base to Scirra in a year and a half, they will be in a tricky position where they feel they need to either make the free version less free, as other companies have mostly unsuccessfully attempted, or to pivot their pricing model in a way that doesn't show weakness, as customers can smell that from a mile away (Scirra has even implied that as being part of the reason they don't do Humble sales). By weakness, I mean implying a concession that they were over-zealous with their price change from C2 to C3.
In that difficult position, I anticipate Scirra will pivot in an insignificant way to attract more subscribers (and even less likely to bring back ones who have "moved on", old C2 users or not). Meanwhile, I predict they will be considering selling the company as YoYo Games did to a private gambling company, or to a larger company who would possibly open source it. Godot for example is free and open source but is sponsored by companies like Mozilla and Microsoft. I doubt the latter scenario (open-source) would happen.
--------
There's not much material value in any of my predictions, so I apologize for getting carried away there.
I do strongly believe that there are good opportunities to benefit everyone, and that they don't need to be delayed, as they do not require much work to implement (a blog post + a couple of additional web pages/sections and form options).
Thank you to anyone who read this.