And this is why been counters make for terrible managers of tech companies
I'm not saying that at all. I'm saying that depending on where you are on your dev timeline, you can't release a game on everything. You will always have to choose. If you had started dev in the past year, it wouldn't make sense to release for previous gen. Because by the time you release it, growth in next gen would have caught up or exceeded previous gen.
Even ignoring the Cyberpunk issue, this happens every generation. Previous gen will eventually no longer be supported, no matter how many owners there are. The first to jump ship to next gen only are usually third party studios. The first party ones have a more vested interest in keeping existing owners happy.
This usually happens when growth in next gen starts exceeding growth in previous gen. Studios want to showcase their work more impressively, and the best way to do that is on new hardware.
Even halfway through last gen, there were probably more PS3/360 consoles out there than PS4/X1. There probably still are. But they ran the numbers and it would still be profitable if they didn't have to support old hardware.
Also being in IT, the dev timeline and roadmap is important. You know what's in store for the future, and you will start shifting your focus towards that. Why would you want to keep supporting old hardware, instead of encouraging people to upgrade so you can sell a more copies of a more impressive product?
Now back to Cyberpunk - dev didn't actually start 8 years ago, that's only when it was announced. Dev only started about 4 years ago. This put CDPR in a tricky position in terms of their dev timeline. They knew that by the time it would be ready, it would be towards the end of the current gen. But they also knew that it would coincide with the next gen, and therefore put a lot of focus into the PC version to make a next gen port easier. Their issue is they tried to kill too many birds.
Rockstar faced exactly the same issue with GTA V. Their solution? Initial release was for PS3/360 only, followed by PS4/X1 remaster a year later, followed by PC a few months after that (their initial plan was to release PC alongside the next gen systems). This allowed for a more focused and polished product.
If CDPR followed Rockstar's approach, I believe they would have been in the clear even if they would have faced some criticism from impatient PC owners.
The point I was making about chasing 10 million instead of 114 million, is if you had to start now (or even the past year or two), you would rather target a faster growing base than a slower one. By the time your product is ready, your target base would be larger, and you have a better product to draw more sales. But depending on where you are on your timeline, you need to make a choice. You can't simply go after the numbers and release it on everything to try and satisfy everyone. You need to focus on doing one thing right. Then you can shift focus to the next. This is what everyone would have learnt from Cyberpunk...