I've only done repairs so far on my one wheel unicycle. Some problems with power MOSFETs overheating, battery short circuits. I've blown 30A fuses accellerating hard fully charged (instant at start) and blown FETs from overheating (running for a while).
Mods are tricky due to the integrated charger in the control board on mine and more importantly the self balancing circuitry being tied to the response and power of the wheel. My thinking is that to up the power a little, better FETS and a couple more (and better) cells might work if I can find space, but due to multiple failures already due to high power, mine may already be pretty well optimised. failures cause face plants as well of course
Rewinding the motor is speciallist though cooling can normally be improved for more power but involves some serious hacking. The motor on mine is basically a stepper motor with 120 steps per rev (fairly complicated hardware) with a phase shifted sine wave drive, (probably pretty standard stuff now), but the extra complexity with a one wheel is that the accelerator is also the closed loop control that keeps you upright. It relies on the response of the electronics and the motor to do so. Increasing the power will mean that a signal results in a greater response which might cause the system to overcompensate and become unstable if the changes are significant (it might shake itself to bits or it might be ok).
Assuming it doesn't shake itself apart and the software allows it, then the problem with running this type of motor faster (top speed) tends to be overcoming inductance. Higher voltage helps this if the rest of the circuitry can take it (More cells in series). There will be tons of advice on this forum on that
if you want more torque at lower speeds for hill climbing, tricks and acceleration then higher current is the way to go (parallel or better cells), again lots of people here can help with that
Good luck - interested to know how you get on
