Well, I'll try to answer as well as I can.
My experience with police officers, and I mean no disrespect, is that if you give officers a tool with all kinds of caveats attached to it—you have to do it this way, that way, make sure the wind isn't blowing, have your back to the wind, make sure you don't have the window open, check the car for spills—and you expect the officer to do in a very rapid time frame, the more likely it is that one step or two steps will be missed, and that is a very serious thing once we come to litigate that case.
Counsel for the defence have an incredibly important role in our society, but they are extraordinarily good at looking at procedures. The minute there's a small deviation, even though it may have no implication whatsoever, they will be litigating that to the end of the earth to try to establish case law. That's fine. It's not a concern for me as a scientist. However, if we can avoid that sort of burden on our courts, I think it would be useful.
That said, it could be useful for individuals who are very conscientious and use it appropriately. I think it could probably add to the arsenal that police have to detect alcohol in people or around people who are driving.