Hmm...Let me try to recap my thinking. You could be right...After all, I've only been brainstorming with myself.

That said, currently, there is a freewheeling diode switched in during braking to get CCM stator current. I am undecided whether this is truly necessary, but it seems like the RMS braking current is somewhat limited without it, in DCM, at least for this half wave SCR configuration. Also, if the control switch is flipped back on while braking, I was planning to wait for the DC stator current to collapse down to zero before applying AC again. (Necessary? I don't know. I figured this would be rather stressful on the toggle to switch this directly.) So, the relay (SPDT, essentially) follows the SPDT control switch, but with some delay and/or condition(s) that needs to be met before switching.
It does add a bit of complexity just to switch in a diode and to (maybe?) save a switch contact. In the current configuration, the toggle switch would only be subject to breaking the motor current and breaking the braking current, but not making the contact for either which would be handled by the relay(s). The DC braking current would be ramped up electronically anyway, and I figured, with a hybrid relay for the 'run' contact, the overall reliability for switches and relays should be pretty good.
As far as safety and robustness, I guess that's another issue. I haven't given it the full DFMEA treatment.
I attached a rather wonky print of the simulation. The relays in question would be U11 and U14, in the middle, above the box labeled "stator". Like I mentioned, it's only a very crude simulation to get a feel for it. Lots of current measurement resistors, and random things that only make the simulation run, but maybe you get the point.
Thanks!