Well, in my view, you can't look at any data surrounding this unless you include data that would include 50/50 pitches. Any data surrounding 3-2 counts might count for something, but not what the percentages are when taking a 50/50 pitch - unless you know how many 50/50 pitches a batter faced in a 3-2 count. You can't count all pitches in a 3-2 count because they are not all 50/50. Sometimes a batter will take a pitch right down the middle on a 3-2 count. Sometimes batters will swing at a pitch clearly out of the strike zone on a 3-2 count, so all ABs cannot be counted in this scenario.
The only way you can do it is to have Pitch FX or something similar to track 3-2 pitches and quantify them into a 50/50 pitch. The problem then is, a pitch is either a strike or a ball. What is a 50/50 pitch anyway? I think we all intuitively understand what it is (I'll know it if I see it), but to quantify it, you have to define it. So, would it be a pitch in which 1/2 of the ball is in the strike zone and 1/2 is out? Would it be a ball that is just barely touching the strike zone?
I hope you get what I'm saying. People ask for data to support this theory, but aside from possibly MLB, there is no data out there to support or deny any of it. Even using pitch data to see exactly where the ball came across the plate, you would then need to define a 50/50 pitch in empirical terms because Pitch FX (or whatever) would only see pitches as a ball or strike - unless otherwise defined.