Can pseudotachylytes be used to infer earthquake source parameters? An example of limitations in the study of exhumed faults