3. As a dev, this is NOT what is currently going on in the current PC space. When we make a PC game, we have a GIANT marketplace & install base that we can sell to. We aren't assessing Steam's potential install base when making a game because we already know the potential install base is over 100+ million (obviously we don't budget according to that figure).
Before we even break this down even further, lets just start off by pointing out the obvious - PC support, EVEN when its a game focused exclusively for PC, such as X-Com 2, still yields wild & unpredictable proper functionality for the wide range of PC specs that currently exist in that ecosystem. When we consider games that have focused primarily on the console version, and then had a PC version also done, we've had significant bad examples this year alone. The fact is, attempting to support the wide variety of setups the PC platform has offered has caused huge, noticeable issues that have become increasingly prevalent. Consoles have offered some more security in this regard, but its not like console development has escaped these issues either. However, console releases tend to become way more 'guaranteed' stable upon release the longer a generation goes on, as long as the developer is leading console development first, and not on PC.
I've shipped several PC games already, and I can tell you just getting the support for various individual components is a development pipeline nightmare when it comes to optimization. Thats why so many individual PC setups slip through the cracks, and we get 'hilarious' youtube videos of issues come launch day calling the developers lazy as a result. Few QA teams can even handle doing all the proper technical testing for these setups. It isn't impossible - in fact, leading on console has helped PC releases, since it gives us a sort of old & weathered benchmark that we can work off of, while throwing in our higher-end render bells & whistles for PC users who can afford to run them.
So now, MS goes "we want to introduce a new closed box platform into the market that is a smaller increment stronger than a full-on console generational step". So, with just that alone, if the APU for the system is not the same as the prior console, we are looking at adding in a new coding development pipeline for that build of the game if its to run natively on the new machine. On top of that, that entire build will need its own branch of QA support. All while we're still building the last Xbox's version.
Not only would we have to hire more people across the board to support it, there is zero guarantee we are primed to make any more money; remember, if the early market is going to be all enthusiasts on these new machines, and there is zero indication in market history that says a casual market will catch on to this, my team is now basically dividing the potential sales we'd have on one ecosystem between two platforms. Except, the act of dividing it actually costs us more money & development time. And there is zero guarantee the new machine will ever have an install base large enough to give us a positive ROI.
Now, here's how the future may look like:
