That may all be true. But
1) Do we know they didn't drink when they got home at 10pm on Sat night? Do we know what they did do? (I may have missed discussion of that.)
2) People who weigh less tend to be more affected by X amount of alcohol than people who weigh more.
3) According to some research, women, regardless of size, may be more affected by alcohol than men and may take longer than 1 hour to process one drink.
4) A person doesn't have to be "falling down drunk" to have his/her memory affected by alcohol. People who are in the middle of an alcohol brown-out or even a complete black-out may appear relatively normal, for example.
5) Didn't the witness say her memory WAS affected by alcohol? To say she and her memory weren't affected in a significant way at the time of the murders, does that mean we have to assume she was just saying she was affected but really wasn't? That she clearly remembered what she saw and heard without "prompting?"
I never found the reported witness testimony to be compelling even before drinking was mentioned. Frankly it sounded coached to me, at least what was reported. And what she supposedly said she saw in the dark in a brief glimpse (athletic but not muscular) didn't make sense to me.
The problem I see is if weak, confusing, or really iffy evidence is presented to support guilt, if jurors don't buy that piece of the evidence it may affect how they view other evidence. And sometimes that is a wise approach. Innocent people ARE convicted based on iffy, coached, and constructed pieces of evidence. Not saying BK is innocent-- I want to see the actual evidence vs press or PCA partial reports of it. But we know innocent people are convicted in the US and sometimes sentenced to death. (In the Innocence Project cases, most wrongful convictions related to flawed eyewitness testimony and government misconduct.) And we know some innocent people have been executed. See, for example,
Cameron Todd Willingham: Wrongfully Convicted and Executed in Texas - Innocence Project
MOO