If I could add to that, if the argument is that there are so many drivers who are very intoxicated, but such seasoned drinkers perhaps that they can escape even this very low standard, the problem is that randomly stopping people isn't going to catch those drivers. You're fishing in a very big pond, and you're not going to catch those drivers without stopping everyone.
Proponents say, “Well, it can be used at a RIDE checkpoint.” There are a few issues with that. That doesn't cure the constitutional infirmity that's there, because we know, and the courts have told us, that police discretion or crown discretion doesn't cure something that's unconstitutional.
If we just employ these random testings at checkpoints, I think there might be a different section 1 analysis going on under that sort of regime. With the amount of extra time it takes to do these tests, even if it's 30 seconds or a minute per driver, it could lead to detentions at these RIDE checkpoints that might not be saved under section 1.
There are some problems with overbreadth, underbreadth, the reliance on discretion. Ultimately, you're giving police a power that we know from past experience will operate disproportionately and probably unconstitutionally.