I'll take this one, since the rest have been replied to. You are once again looking at a misleading picture. All of these tests are done with the same tactic and instructions, but it's obvious that an "extremely high tempo, attacking, counter and counter-pressing" tactic will be punished the most by lacking pace and acceleration. That doesn't mean that those are the most important stats, it simply means that they are the most important for this specific tactic to work. Furthermore, do you really think that off the ball, finishing and decisions don't matter either? You must, since they score nearly identical to height. This is a good example of your data only being as good as your methodology. What would happen if we increase all attributes by 4? What if we played a different tactic? What if we played in a different (real) league? See, we must all make these contextual choices in building our team. There's no need to strive for a universal "this is the best way", in large part because it's never a choice between the same CB with 4 less attribute points. Sebastián Coates is one of the best defenders in the game this year and he is not fast at all. He has a great spread of defensive attributes, but can obviously be made to look bad if you play with a high line, intense pressing tactic.
Now, to jump back to the other video, the reason a tiny centre-back can win the most headers is partly fueled by him needing to use his head for crosses a taller player would volley. We saw in the giant player videos that height absolutely influences what a player chooses to do, and this would inflate such numbers. Most of what you've posted falls under unrealistic input, unrealistic output. That's not to say those things aren't fun, but they cannot be used to quantify what attributes are most important.