The motor is NOT adjustable. The mounting surface on the left side is machined onto the frame such that the motor is as much as it can be guaranteed, to be true to the mounting axis of the swingarm, which in turn as much as can, guarantees that a user can then align the wheel sprocket using its adjusters. The position of the belt on the sprocket is probably a red herring. But by all means, keep wasting your time trying to strain the motor against its mounts. At worst, you'll prematurely destroy your $1k-2k motor, assuming you can get a replacement at all since this wouldn't be covered under warranty.
First thing I would do is completely slacken the mounts on the right side and back off the adjusters so that the motor is held only by the left mounts, which should be more accurately machined relative to the swingarm that you'd be able to measure. Then, mess with the wheel alignment and see if you can get better results with the belt noise -- NOTE, don't heavily load the motor with the right side unsupported. I can pretty much guarantee the motor doesn't have the torque available to tear itself off the left mount, but the added movement from lack of support on the right side wouldn't do it any favors. Also, for this test, the belt tension would have to be set very low. You can try to set it to the bottom of spec, but you'd want to put some kind of feeler against the right side of the motor to make sure that as you tensioned the belt, the motor wasn't twisting. If the motor starts to move, stop applying tension and just try to get the alignment right.
If that fixes the alignment, that tells you that someone messed with the right-side mounting bolts and racked the motor. If, however, you still can't get the alignment better and feel you need to rotate the motor, you can adjust it ONLY by adding shims under the bolts on the left side. Again, the adjustment of the mounts on the right side are there solely to facilitate mounting the motor from the sides without stressing it or the frame when doing so.
And just to make sure this is clear (even though I've said it several times now), "better alignment" does not mean the belt is riding in the middle or even off the flange. You consistently describe the belt as riding up the flange, even if you don't understand that definition. When the alignment is correct, the belt might very well still ride against the flange, but you should not see any part of the belt being lifted by it as it engages. Don't let its position on either of the sprockets be some kind of red herring to you.