if your only issue is with the level of the BAC - then argue that point. even with a BAC at .10 or .12, you would still need some form of pro-active policing to catch people driving at that higher level. otherwise, what you are basically saying is that the only function for the police here is to clean up in the aftermath of DUI related accidents - the innocent driver is already dead or maimed for life.
are you in favor of any deterrent measures at all? DUI is not like speeding - you can't aim a radar gun at a car and determine if the driver is DUI, to develop the probable cause to stop them. what's available? checkpoints. OK, you don't like those. how about we have task forces of unmarked cars that follow people as they leave local bars to observe their driving habits. can we do that, or does the right to privacy extend to not having a cop observe me as I drive? after all, I might be innocent, why are they following me?
We don't need a subset for DWI, cell phone usage, radio tuning, ogling members of the same or opposite sex on the sidewalk, so on and so forth.
The issue isn't the numerical value of BAC. Rather, the issue is a willingness for MADD et al. to ignore the fact that the most dangerous drivers are clearly visibly impaired. If someone is not visibly impaired, odds are pretty good that he's not one of the most dangerous drivers (or, if he is, it has nothing to do with alcohol).
If a particular person is capable of driving his vehicle sufficiently smoothly and safely with a 0.11BAC that an observer wouldn't be able to tell that he was drunk, I don't care if he 'slips through the sytem'. The people I want off the road are the ones who can't control their vehicles safely and effectively. And observation is probably a better predictor of that than BAC.