Yup, and the "preprint" (which doesn't have a number of controls in the process) leaves a lot to be desired, so the "real" paper will presumably have some of th

Yup, and the "preprint" (which doesn't have a number of controls in the process) leaves a lot to be desired, so the "real" paper will presumably have some of this worked out. I expect things like the cooling rate (which affects crystal growth) and oxidation will both have variability in them.