Seriously, I would rather they did a top down 2D game, with a custom engine and did it really well.
I find the discussion here is just bizarre. What problem would a custom engine solve that would justify the expenses? As a software developer you are paid to solve problems and not to write code. And as a game developer you are tasked with, you know, creating games and not developing engines and tools. You want to use off-the-shelf software whenever possible unless you can make a strong case for doing the development yourself (e.g. prohibitive costs, market expansion, software not not available or modification too costly, etc.). What is this absurd fascination with reinventing the wheel whenever possible? Do you also build your own hammer whenever you need to put a nail into the wall?
Making a top down 2D engine, with the same functionality would take a week. Heck there are ones they could take and use right now.
The time consuming part of coding an RPG, is the RPG part - interacting with creatures and objects, the rules that govern those interactions and the enormous amount of scripting. They have to do all those things in a 3D engine, just as they would have to do them in a 2D engine.
Of course they aren't going to change track, but top down 2D, is what I assumed they were doing from the start. Old school is 2D.