They did a really good job on it.
I saw a draft of the document back in May, and there was a whole, whole lot to like about the document. It was well written, very thorough and covered all the bases that needed to be covered, but as a result, it was not light reading -- unavoidable, comes with the territory.
The draft did not talk about using recorded digital flight data like is on the Garmin G3X Touch and also on the Dynon. (I think the G3X data is a little more complete, although the Dynon is recorded two times per second instead of once). I've been thinking that when I put the constant speed prop on the RV-9A, I'll probably fly performance tests using the digital data, and I can probably do it much faster than with manually recorded data, especially with the autopilot flying. Then again, I've done FOQA development in my career and am comfortable with those processes.
I will admit, though, that when it comes to using performance numbers in the real world, I use such fat safety margins that precise numbers don't come into play -- and I dipstick the tanks all the time. I've found that on really long X-C, sometimes the destination airport gas is unavailable, so I try to land with two hours of gas so I can go somewhere else if I need to. It's saved me at least once.