I have not suggested it's not complex, I'm a Software Engineer so know very well.
However I have zero experience in gaming but wondered what after all the hassle of upgrading Red Engine to produce Cyberpunk (and look amazing) they'd jump to UE5, cost is clearly the obvious answer but I thought I'd asl.
Any expertise to share?
Yeah this wasn't directed at you specifically, you asked a valid question.
Cost isn't the immediate answer, although budget always matters in the end.
In the DF video about the Witcher 4 Tech Demo, as a collaboration between Epic and CDPR, one of their tech directors cuts into that topic briefly; worth watching to get a general impression.
But long story short, an engine is far more than "eventual output". People often mistake an engine for primarily graphics or physics. They look at some great looking game, or some cool physics in a game and go "wow, this engine is impressive"; and it's not wrong, but something like this is 'just' some of the fundamentals an engine does.
Regardless of cutting edge graphics or similar, even more important is the quality of the tools. Tools decide how efficiently you can create a game, and downstream of efficiency you have headcount, cost, amount of content etc.
Even if you keep headcount and budget the same, with improved tools you will be able to create more content; more quests, more NPCs, more dialogue etc.
A very basic example is motion capture. If motion capture relies on big, expensive setups, suits, markers, expensive license fees, and manual cleanup, this directly impacts how many motion captured dialogue scenes you can put into your game.
For AAA, there is a quality bar; if, say, a quest is at a certain technical quality (focusing on animations and voice acting primarily), and a typical quest takes 5-10 people a month to make (motion capture, actor, programmer, vfx, sound, cut scene direction, storyboarding, writing...), this will directly limit how many quests you can put in a game; after all, quests that will be subpar in comparison might simply be cut.
Now with the evolution of motion capture, also demonstrated at Unreal Fest, this becomes more accessible to devs over time.
Smaller devs get access, and bigger devs can get this quest done in half the time, doubling the amount of quests you can put in the game or letting the devs focus on other parts of the game.
This doesn't go for just things like motion capture and automatic lip sync, but _anything_ you can imagine. A good quest tool will allow more quests to be integrated. A good quest tool will have a fancy event system that lets you make more interactive quests more easily, reducing programming overhead. The same goes for world generation; do you want to hand-place each grass cluster or maybe there is a way to do this more efficiently via Unreal PCG? It's not that hand-placing everything is impossible. But it takes time. Time that could be spent elsewhere.
One example the tech director mentions in the CDPR video is how while everything in their engine is multithreaded (not the case in Unreal), it's difficult to work with. So the output can have the potential to be great, and to not be held back by technical bottlenecks, but iteration time suffers, and that will directly impact scope and quality of content. Remember the cut content of Cyberpunk and how it differed from earlier trailers? Some amount of setbacks and cut content is completely normal, but while the tech might have been great, the result in terms of game design, mechanics, content etc. would have looked different in Unreal. How different? Better, worse, in what ways? I can not say.
But these are, outside of the financial reality, the main factors to consider.
EDIT:
Man, I just finished my response and the dude is banned? What happened?
), there are games with stutters & bad frame-pacing too, but people tend to lump them altogether in order to keep raising their voices to feed into this demand of "60/higher FPS or die!!"
