Maybe you should read a little more about the fella that ELK posted. He makes the opposite case you just did, her name is sarah and she had a glass of wine with dinner and then finished another after her dinner, one more quick drink and she noticed the time and had to go.
Then she gets pulled over down the block and the time it takes to get to the cop shop and take the BAC level it is actually MORE than what it was when she was arrested.
Seems I made that case a while back in this thread didn't I?
The truth is there is no way to know all the factors that need to be known to figure such a thing. Thus there is no way to apply a law equally to all people. That I think is one of the factors that justifies .08.
Such things can be pretty well estimated. My understanding is that a rapidly-rising BAC will have a much greater effect on a person's abilities than would a stable BAC, and I would see nothing wrong with declaring that a person who has consumed enough alcohol recently to raise his BAC measurably above 0.10 (by whatever the +/- uncertainty is on the instrument used). If someone who is driving very poorly (so poorly that he'd be deemed unsafe to drive even if sober) and blows a 0.07, then fifteen minutes later blows a 0.13, I would argue such a person should be prosecuted for DUI even though--at the time he was driving--his BAC was below 0.10, since such a person would have consumed enough alcohol before driving to raise his BAC well above that limit.
The biggest argument I've heard in support of 0.08 is that it's necessary to prevent the really dangerous drivers from getting off on 'technicalities'. I'd suggest that there are better ways than the 0.08 standard to identify the dangerous drivers.