The police are likely right that it's too early to tell whether the shift to .05 from .08 has saved lives.
Labour's Sue Morony very correctly points out that very few crashes involve drivers between the .05 and .08 limits. There was never much chance that the change would have passed muster if based only on deaths in that cohort - there's just not that much increased road accident risk between .05 and .08.
The chances of the policy's success rather depended on whether .05 meant fewer drivers got to .10 or higher. If people make bad decisions about having another one or two after hitting .06, and don't get to .06 under .05, then the policy could save lives. Whether it passes cost-benefit is another thing. But it'll be interesting to see the police's numbers when they come out. I hope that they properly adjust for prior trends in drink drive rates when they do so.