I may be picturing your description incorrectly, but the rear ring on that motor adapter does exceed the OD on the 98mm case.
The piece that exceeds it is the outer tube sleeve, as opposed to a piece of plywood.
So, what you have in the motor transferring thrust to the adapter body, then the adapter body transferring it to the motor mount.
Basically, the thrust ring on the motor transfers its thrust (mostly) to the motor mount tube in the adapter. That tube, via shear load, transfers it to the rings. Those rings, via shear, transfer it to the 98mm outer tube of the adapter. That tube transfers it via shear to the thrust ring (a sleeve of tube) on the adapter.
With construction like that, it's tricky to get a full bonding surface on the tube-to-tube joint.
Instead, I prefer when the piece that takes the load from the motor's thrust ring and transfers it to the rear of the rocket is one solid piece of plywood.
if requested, Dave will drill a couple holes through the rear ring and 98mm tube and run at least a 1/4" dowel rod through them, then sand them smooth.
That puts the dowel pins in a position to help bear the load, and isn't a bad idea. But, by redesigning to use a single piece of plywood, that's no longer necessary, and assembly becomes
much easier.
I've seen an adapter built similarly to the one you show fail, resulting in a J570 shooting up through a rocket. We examined the adapter afterwards, and there was a complete glue joint that had failed; the rest of the adapter was built with the same batch (including mix) of epoxy, and it appeared to be rock-solid.
Needless to say, this changed how I look at adapters....
-Kevin