Then perhaps an outer o-ring (encircling all the bolts) would have been a better choice, if mating surface alignment is very important.
I'd think because it's a rubber belt being driven, there'd be decent tolerance for shaft skew taken up by the belt flexibility. It also appears the belt can handle coaxial slippage on the sprocket (self-positioning) if it's narrower than the sprocket between those side guards that prevent a belt from slipping right off the end of the sprocket. ..
Hey, while we're on this face mating topic...
These pictures show lots of corrosion on the faces. Do you think pressure of metal flaking and expansion with road salts continuously being added is one of the factors causing the bolt shearing? Are the faces being pushed apart by corrosion once it takes hold between the surfaces?
If I find one broken bolt, the motor is going to have to come off. Then I'd be re-polishing the mating surfaces and slather thin coating of anti-seize over it before reassembly.
And in all cases, some kind of sealant around the seam once its bolted back up.