Either way, they’d be far better off trying to start with something that already works than starting over from scratch - see the 30 years of development/test/debug comments above 
I’d wager that between the companies they own there’s a huge amount of experimental code in the various engines that has a use case but never saw the light of day with a release.
See this is exactly why studios make their own engines.
Idtech6 is made for an arena FPS game. It have perfect aim, it have high Res textures, it have great spacial movement. It doesn’t have anything that will help an open world game, nothing for a mellee combat and absolutely nothing for 3th person gameplay…
Doesn’t nintendo make like one engine per game almost?
Maybe way back when… But nowadays, I’d imagine they’d craft a base engine, for each relevant gameplay style
I mean obviously they reuse the breath of the wild engine for breath of the wild 2. But im not sure they use it for anything beyond that.
Haven’t touched game development in many years, I would say from a marketing/creative side a game studio would want as much “control” and least amount of money going to another developers’ engine–this leads to innovation through risk.
When I was younger my attempt at working on a game engine designed for open-worlds was a pain in the rear, OpenGL was a requirement as the base project started around OS X/Linux… dealing with frame buffers/caching pre-large RAM GPUs wasn’t fun, if I recall texture compression on GPUs didn’t happen until the last few generations of GPUs. R&D on game engines requires a ton of long term investment.
Can you elaborate here? I would think that a AAA could just extend the frame work. UE advertises the engine as being fully exposed in terms of the source code.