The work hardening thing got me thinking... Work hardening involves plastic deformation, IE: at least the yield stress. Deformation in Ti involves lots of twinning (rather than dislocation motion) to get any deformation along the c-axis (HCP crystal) and I found something that says that twinning is more favourable at lower temperature, and I'm fairly sure I remember twinning being more effective at work hardening (than dislocations)..so..it seems quite possible.. Geek moment over.