I don't think "just" the sensor board affects the alignment. It could be affected by the rotor and or the stator can itself imo. I've found that the way the Trinity motors are designed, when you adjust timing, you can slightly shift the rotor off center. This could affect the alignment of the rotor to the sensors. That said, it doesn't seem to have any real impact on performance that I can tell.

I do think how the sensors are placed on the board has an impact on performance though. Of course the proper placement and alignment has a hand in accuracy, at least imo anyways.

Different rotor / shimming of the rotor relative to the sensorboard will cause different readings. Also the Skyrc type motor analysers are limited in their ability to accurate record sensor angles compared to a motolyser etc

Sensor board is not centered, play a bit with the small amount of mounting play you have on the board.

I'm actually surprised you could remove the sensor board and lace it back and have the accuracy the same. I mean I think you could fool with it and get it close, but the same is surprising. I spoke to trinity about this with their units and while they acknowledged it (the mounting play in the timing part which also contains the bearing) they seemed to dismiss it as any sort of issue.

I took apart a fantom motor to compare and they way they do it seems a lot better, bearing doesn't move (no play) when timing is adjusted. I think the tekin motors were like that too.

The sensors as soldered on the boards, I think fantom was the best placed followed by tekin then trinity but those two were pretty close.