I have refrained from commenting in this thread since its inception, but reading "UE5 will have to be applied to everything to aid realism, not just the terrain?"clearly forced me to drop a few thoughts, hoping to add some clarity to those who are not familiar with software and specifically game development. Apologies if I repeat thoughts from many others.
A game engine is to a game what Android or iOS are to a phone, Windows or Mac OS are to a computer. It's the foundation that underpins all the mechanics in the part of the game that runs on our computer. Also, it has nothing to do with the part of the game that runs on MA's servers, though any server/network induced loot lag will affect it. In other words, i's not a simple graphic effect, which you can apply here and not there. Nor can it be applied to some planets but not others.
Operating systems are inevitably linked to the hardware of the era in which they are developed. That's the reason why they evolve over time. Taking into account the speed of change in hardware and software in the last 10 years, especially in the gaming industry, the game is the equivalent of a program developed for Windows 95. Sure it still runs but 1) it might not even run for much longer; 2) it's almost impossible to maintain and develop; 3) it looks and feel from the late 90s.
We have grown to consider normal for our avatars to fall through the floor of a shop or get stuck in a tree or a wall, or worse plummet in the underground ocean. That is the work of a game engine, which makes sure that the laws of physics are never broken. Game developers should never write a single line of code about that, after they have described what an object is. Same as a phone app developer doesn't have to take care of the code to make the phone connect to 4G networks.
I have also read comments about the "potential" of the existing game engine. That is frankly utterly ludicrous and delusional. We all know that EU runs on CryEngine 3, but not many seem to be aware that Amazon bought CryEngine 5 a few years ago, to use it in their future AWS Lumberyard cloud game engine. They actually had to rewrite it from scratch to make it suitable for their plans, and even so, it hasn't been a success, to the point that just few months ago they changed strategy and open sourced it, hoping to get some traction.
Any gaming company starting from scratch in 2021, would have Unreal Engine (either 4 or much more likely 5) firmly at the top of their game engine shortlist. And the list is short indeed, I would add Unity, and possibly a few more sector specific ones, but that's about it.
Rather than looking at any technical transition problems from past experience from 15 years ago, we should all be delighted that MA has taken this step. For many investing time and money in this game, the biggest question has always been "is this game being developed and going to be around at all in 10-20 years"?
Looking at the current state of the game, one could be forgiven for thinking that MA were not really developing EU anymore but just looking to use it as a cash cow to squeeze money for as long as possible and then shut it down when it could not run anymore on, say, Windows 12.
By taking this step MA is announcing that EU not only is here to stay, but that they have plans to gear up its development both from a technical, and probably even more important, from a marketing point of view. And that is something that the entire player base has been screaming for years.