Läsprov: Concerning the weakest coherent formalization of methodological skepticism as a Bayesian updater