Well now. I would be able to send it back if I hadn't taken it apart to paint it when I got it. Also I've then it apart again to fault find. So it looks like I'll take one for the team and do a full autopsy.
Micro board is still fine, still connects to pc via program card. This board houses the FET drive ICs and the BEC regulators, all still working fine.
Top board is the FET board which has 24 mosfets these are alpha omega AON6512 in a DQFN 5x6 package. This where I suspect the quality problem may be. I'm not a fan of alpha omega semiconductors. In a high current application such as this personally I would have specced an international rectifier part instead. But hey ho... The issue is as there are 4 fets in parallel for switching phase you only need one to fail for the whole stack to go.
So lesson.
Don't paint your speed controller
Matt