My own opinion is that the risk outweighs the benefit of actually stopping the engine of a single at any point.
But we're not talking about "actually stopping the engine", are we? We're talking about pulling the mixture to ICO. The prop won't stop, unless you slow the aircraft down to a speed that allows it to do so. The engine is still turning.
Why is there a better chance of the engine producing power again after the throttle is pulled and restored than after the mixture is pulled and restored? One line of thought says that induction icing is far more likely if you use the throttle to simulate zero thrust than the mixture.