Oops, missed this when it was fresh. I agree that the coarse hard particles could likely cause this type of sub-surface damage pretty easily. The key thing is to look at it as Victor described - the impact occurring in line with the apex as the edge is brought into the particle in question. This is a tiny area of impact, so it wouldn't take a whole lot to get to that 100ksi number.

It isn't probably going to occur in the other plane along the flat area of the bevel - which makes sense, as all the damage seems to occur only at the edge. The impact could be enough to fatigue that bit of steel that gets driven back to the point that it weakens and gets ripped out later after more honing and possibly flexing a bit.

Another thing to consider - using the diamond plate in the opposite direction (edge trailing rather than edge leading) might well prevent this damage, after which one could switch to edge leading for the 1k and up stones. I am quite keen to try edge trailing strokes for the whole job until the very end as well after seeing some of Todd's latest images and ideas at scienceofsharp.