Thanks - for some odd reason it looked like an O-ring at first, which made no sense at all.
That sounds like a reasonable approach. A knowledgeable friend is willing to stop by and check it (and my grinding technique) out in a week or two, so I'll probably wait until then to dig into the spindle. It's been suggested to me that the runout characteristics (max and min just about exactly 180° apart) indicate that the spindle may be slightly bent. If that's true, working on the bearings probably won't help, but then dressing the wheel sufficiently should eliminate any eccentricity that would cause scalloping.
Someone else though that the table rack or gear may have some sort of problem that is contributing to or causing the scalloping. I'll check that out tomorrow night after work. At least the rack would (probably) be cheaper than new bearings, but it won't do any thing to reduce the run out.