I think that the problem is far less than commonly imagined. I have run magnets on a test rig just held on by the magnetic attraction. The field holds them pretty well but at high speed they would move out radially.
Ones that I have stuck on with Loctite structural adhesive ( Multibond 330 here in UK) have never moved. I have some machines with the magnets stuck on with this and potted in polyester for weather protection, none have ever moved but I suspect the polyester block is not really stuck to the steel disc as rust seems to be forming under it on an early machine.
With a decent epoxy and proper surface preparation I don't think they will move. If you cast them in an epoxy block then they can only move together. To be of any real benefit you need to key the whole block so that it can't go off centre. This is what the stainless band does very well. In your case it sounds as though you have nothing to prevent the whole block moving off centre so you are no better off than epoxy alone.
If it already cast you could drill a few holes through the resin block into the steel discs, tap the discs and fit a few screws. These need only be small and even steel screws will not affect the field if you go somewhere between the magnets. Alternatively perhaps even simpler is to drill the holes and fit roll pins. With 4 1/8" roll pins the magnet/resin block will never move even if it comes unstuck.
I think much of this worry has come from cases where magnets have come off after a failure when the magnets have rubbed the stator. These magnets are demagnetised and have no natural tendency to hold to the disc and are a very different issue. As they are scrap it makes little difference if they come off. Just don't fry the stator and this is not going to be a practical situation.
Flux