I mean, most current engines support essentially everything Gamebryo does, but better, with fewer bugs and better performance.
You are right, if you compare Unreal Engine and Unity to
Gamebryo, they most likely provide everything that it does. However i'm being pedantic here because "Gamebryo" isn't what Bethesda uses, Gamebryo is a general purpose engine (or actually, an "engine creation kit" or framework of libraries) that can be used to build other more specialized engines - it does not provide much in terms of ready made tools. I saw an evaluation version briefly back in the mid-2000s and it had very barebones tools and i'm sure NetImmerse, the version that Bethesda's engine was based on when they started working on it during Morrowind's development, was even more basic in terms of tooling.
However you are right in the sense that Gamebryo does provide everything Unreal and most likely Unity do and just like these two, Gamebryo had been used as a basis for (according to their site) more than 400 games. Some can be seen in
this video from 2014 and in
this wikipedia page.
However what Gamebryo doesn't provide is anything associated with Bethesda's engine and what provides the actual game mechanics and most of their support systems. Those are made by Bethesda. Some examples, like how NPCs and PCs have stats, etc and the overall world interactivity were mentioned already by others.
However one thing that is important and these rely on is what is commonly called the world database - essentially how an engine tracks what is in the game world, how these items/entities/etc are placed, removed, etc. Engines differ a lot on how they handle this and Bethesda's engine is quite unique here.
Unreal Engine? The Outer Worlds has essentially the same gameplay as Fallout 3/4/ETC, with nothing missing. The worlds in the game are smaller, true, but this is a limitation of the game, not the engine. There are
several open world games created on it.
Bethesda's engine has inherent support for seamless open worlds (what people often associate with "open world" nowadays, even though something like Fallout 1 is also an open world game despite not having a seamless world). Unreal Engine is not optimized for seamless world games even though you can create them. In Unreal Engine the support for seamless worlds was added as an ad-hoc solution on top of the existing level-based approach that the engine had ever since Unreal Engine 1 and relies on manually creating separate "sub levels" which will then be placed on top of a master level that the level designer has to explicitly script and/or create streaming volumes to instruct the engine when to load/unload them.
On the other hand Bethesda's engine, even in the original Morrowind version, relies on a grid system that loads and unloads individual cells dynamically as the player moves through the world automatically - there is no need for the designers to bother with setting up streaming (outside references but that is also something UE games need to do anyway).
Unity is a bit of a mess, but certainly no worse than Bethesda's. There are also several games with a gameplay that's very similar to the TES series, with open world, interactive scenarios; like Rust, Outer Wilds and Subnautica.
Cryengine is a bit harder to work with, but it also has a
good list of open world first-person games, and is infinitely more stable than Gamebryo.
It isn't just having a seamless open world, but also having persistent items, NPCs states, etc... others already mentioned those. In theory it isn't hard to implement those, but an engine's design can make it
very hard to do it and many engines - especially general purpose engines that often focus more on static visual fidelity - have restrictions in place that do exactly that.
I do not remember where, but i remember reading a Morrowind postmortem where a Bethesda developer mentioned that they designed their engine from the ground up to be able to handle big persistent worlds. Them using NetImmerse (Gamebryo), which especially at the time was essentially a collection of libraries, helped to do that but using something like Unreal or Unity that comes as a complete package with its own ideas of how things should be done (which might not be bad for most of the games made with those engines) can actually work against this.
But it wouldn't be more work than trying to make their engine even remotely functional, I bet.
In general people overestimate the issues Bethesda's engines have. Now, i do not know *why* they are not solving them, but issues like framerate, animations or whatever are really things that can be solved. My guess isn't that their engine isn't capable of having those solved, but that their management isn't giving much of a priority to having these solved. As an example, AFAIK Skyrim VR runs at high framerate, which is essential for VR, which shows that they *can* make their engine run above 60fps, given enough incentive (and, according to a friend who has a VR headset, the game is still very shoddy which means that they didn't even have to spend much time doing it).
Other engines were not built or adapted from the start to be so modular. That way, they would have to either invest a ton of time to try to make the engine more accessible and easier to create mods (which is an extra job and can still make the engine more unstable in the end)
Yes, it is a very big job and chances are they'll end up with a worse product than what they originally had. Making Unreal (or most other general purpose engines) do what Bethesda's engine does means changing things at a fundamental level that would affect every single part of the engine itself, except perhaps the renderer. But the renderer is only a tiny part of an engine anyway (the size of a renderer in a game engine is something that many gamers greatly overestimate - and also something that marketers take into advantage - whenever you hear about a "new engine" or even new engine version, 99% of the time it is an existing engine with its renderer changed to look different from the previous one).
or they would have to abandon the idea of having games that are simple to create mods to, and since they've already cultivated a gigantic community used to creating content for their games, that would be suicide.
And it isn't just the mods, Bethesda themselves are using the same tools and having tools that are easy to work with also affects Bethesda's own designers and performance. The tools being easy to use was also something that Obsidian designers mentioned as a positive for the engine.
So there isn't really a good reason for Bethesda to change engines and use something like Unreal or Unity or CryEngine. They'd have to do almost as much work as they'd do to make a new one from scratch - but making a new one from scratch is a waste of time since they can fix their existing engine (that they also have full control over and have shown they can fix). The only reason they release game with issues is that their management simply does not allocate enough time (which implies they do not consider a priority) for their programmers to fix these issues.
(of course, their management could also decide for reasons only understood by managers - and perhaps marketers - to switch engines, but that wouldn't be a technical decision and that is another topic... i'm just pointing this out because i do not want to imply that Bethesda will never change their engine, but that the engine itself wont be the reason if they do)
