Noticed that the v1 variable precedence docs list facts discovered as having a lower precedence than inventory variables. It is in reality the other way around. The v2 section gets this right.