Couple of things I notice: the FETs are conducting though their plastic casings rather than their drains where the heat can come out the easiest. Junction to case rating is 1.3C/W, its probably 1/5 that for junction to drain pad, then they still need to get through the thermal pad.
It’ll be drawing about 12-15mA of switch current at 10kHz, comfortably under the 30mA rating of the DRV8302. @Jasey could you measure the gate resistors with and Ohm meter?
It has 8x 100uF 63V capacitors. The best 100uF 63V smt capacitor on digikey is this: https://www.digikey.com/product-detail/en/united-chemi-con/EMVY630ARA101MKE0S/565-2503-1-ND/757664
190mOhm, so 8 of them could have an impedance of as little as 24mOhm, and as much as 50mOhm. That is high compared with a 60A VESC4, which usually has 3x ~500uF 20mOhm capacitors. I don’t know if a standard VESC4 is overkill or if this is undersized.
The layout is very good from a noise perspective. There isn’t much room for inductance there, I bet the high frequency switching noise is very small. This is the biggest threat to electronics in a system. It must be small enough they don’t need large electrolytics to filter it out. The board is also well designed for cost. It’s a shame the bottom has so few components on it, its just the transistors and a few resistors and nothing else but a logo. They could have made this board an inch smaller than it is.
I bet it can handle 110A motor continuous at 48V. But I doubt it can do 200A into the motor for more than 20 sec. Not sure what the word “continuous” means anymore.
Edit it says 800A peak! No chance it can do that. Not enough fets. I bet that means it has 100uOhm sensors.