Just to be sure: after many years of engine development with a mediocre demo, you want to build everything in a new engine that you're not even sure is suitable?

If their goal is to make the original idea of XStoryPlayer where you have a town and a few mission things but open to modding then UE5 can handle it no problems.

Animations can be blended and even data driven with control rigs,Ik, so during runtime make a hand follow something.
Particle systems with CPU particles to cling to surfaces...... and im pretty sure they can even trigger hit responses on collided actors.
Ai using behaviour trees and blackboards can do some really nice things and the navigation is really good.
Environmental Queries to allow NPC's to take cover... wait wrong kinda game.
Soft body physics for the bouncy/squishy bits.
Natural Language Processing can be made to allow custom sentences or use UI choice elements for static choices.
Shaders and dynamic materials that can be edited during runtime.
Cloth simulation which i havent played around with yet but im assuming you can edit its states.
Modding with either the editor or make a custom solution either ingame or out.


Nanite and Lumen which gets a bad rep if you dont use them correctly but can look and run great or just ignore both and use the older rendering methods.
MegaLights, i threw 10k dynamic shadow casting lights into a scene as a test and it dropped me to 40 fps, used to add 5 and have that many frames.

XMoon has made a pretty good engine considering they did it themselves and UE4 was no where near the level needed back then but UE5 has seen some really good improvements.
Also making your own engine means you spend more time working on the engine instead of the game.
I would love to see XStoryPlayer brought to Unreal and let them focus on gameplay but it would still take a fair bit of time to get used to the workflow and build all the frameworks.