hmm, I will be "halving" the length of the current path too, so I may be putting more current into each leg for a given bus voltage, which might reclaim some torque. Dunno, gonna have to make a center tap and do some experiments with a battery and a scale and an ammeter
Edit, and see if I can find a combination of hall effect readings from the existing hall sensors that coincides with when the poles should be energized (120 degrees apart, leading pole alignment by 120 degrees)
I would still use something like an atmega to control it, easier for me to figure out/modify than a collection of 555s
(easier for me to add a throttle too also)