If it is a timing issue, you could see it by using an oscilloscope on a phase and it's associated hall signal; they should be aligned (although I don't know the exact alignment, there are a number of controller development threads here with good scope images of what you should see), and the other phases and their halls should be aligned exactly the same timing from each other. If the hall signals are not all the same, it will cause the controller to fire those phases incorrectly, which will cause the motor to spin incorrectly.
There are motors like the Revolt that have poor hall positioning that create either poor timing or poor signal quality; if your motor is like this, then it can be corrected, but will require determining the problem first.
Another potential issue is the way the motor is designed, electrically/magnetically. When a motor is spun mechanically from an external source, it will create a specific waveform on the phases. At least sometimes, if the driving waveform is sufficiently different from this waveform, then the motor may not behave as expected under some driven conditions. So...some motors may work better on a trapezoidal waveform controller than they do on a sinusoidal waveform controller, and vice-versa. Which ones and how much better, I don't know--there have been discussions about it here, but not a great deal of well-documented experimental testing.