I'm looking at making a new printer. I'm thinking a core XY similar to a Voron, but I would be making it from scratch. I'm looking for something I can make reliable and accurate. I want to print PLA, ABS, TPU and more. I have a bunch of parts now that I would stick to.
235mm heated bed
Revo hot end
Nema 17 motors.
BTT E3 mini, although I could use my SKR3 instead.
The easiest would be a bed slinger, but I am open to a Trident style. I like unique and challenging things.
What new features should I include? What should I avoid?
Why not just build a Voron? Other than your bed size, nothing else you listed wouldn't work on a Voron build. If you want to make it your own, the design is fully open source so you can modify to your heart's content. There are also a ton of "unofficial" and "official" mods.
It's worth saying that CoreXY != bed slinger, so...
I design my own printers and like the challenge. Vorons are expensive and have a ton of parts. I like unique stuff. Ive made my own delta printer and conveyor belt printer. The limitations of those leave me wanting a more traditional style.
If cost is your design constraint, and you want a CoreXY design, think about what drives the Voron BOM cost and optimize around that. Without thinking about it too much, the thing that jumps to mind are the 4x (2,4) or 3x (trident) z motors. Reducing beyond 3 will not allow you to get the gantry mechanically in plane with the bed, but that didn't stop Bambu labs from using a single motor on the z-axis for the x1 and p1. Ditching the cable chains for a CAN or USB toolhead would also probably save some cost out of the gate (fewer wires + you won't have to buy the cable chains).
Also consider what your design goals are. In the case of single vs 3/4 z motors, you're trading initial fiddling with cost. A single z motor is going to require more fiddling to get right, but it does save on BOM cost.