The point here isn't linking xylitol to diabetes, but (to me quite surprisingly) links ingested xylitol with serum levels - and then serum levels with clotting.

Showing that dietary xylitol is absorbable in humans is going to need some replication; this is the first I've seen this.