Tautologies are useful to detect errors in reasoning, but not necessarily in actual policy.

The problem with tautologies is that usually the context where they are useful is very small.

As an example, lets use Equation of Exchange and Cambridge Equation. Defining "Money" (M in the equation) has become nearly impossible when there are so many liquid assets and instruments that act money like. Central banks can't reliably measure or control a single "quantity of money" in a way that reliably links to GDP. And because it's not causal model, just an equation, it does not explain anything interesting. All interesting questions are causal.