Is this a trick question? You know as well as I do there will be no consensus on who the "8th ranked" player is. One guru's board might have a player eighth, another board would have the same player 3rd, another would have him 11th. So what is he? He is whatever he is chosen.
This is exactly my point. Showing that higher picks tend to be more successful than lower picks tells us nothing about whether or not 'reaching' is a bad move because whoever we take at #3 will be the #3 pick. Those statistics don't take into account the 'true rank' of a player prior to the draft, they take into account only the player's actual draft slot, and no matter who we take, that player will definitely be the #3 pick.
And it makes sense that the numbers would show a bit of a correlation. The problem is that those numbers show averages. So obvious, clear-cut #1 picks like Bryce Harper get thrown in with years without an obvious #1 pick or generational talent, so Bryce Harper and Luke Hochevar get averaged together. The fact that there sometimes are guys at the top of a draft who are clearly more talented and more likely to make it than other guys skew the data. That doesn't mean that on an average year the #1 or #2 pick will probably be better than the #3 pick, or that the #3 pick will probably be better than the #5 pick. It simply means that over time, there will probably be a player in there taken a little higher than skews the data.
Plus most of this data groups picks together, like nsacpi's link, and my guess is that's because using data that compares each slot individually does not produce a large enough sample to be statistically significant. I could take a smaller sample and say that in 2014, #3 pick Rodon appears to be better than #1 Aiken or #2 Kolek and that #4 Schwarber appears to be the best of the bunch. In 2013, #2 Kris Bryant is far and away better than #1 Appel. In 2010, #3 Manny Machado is definitely better than #2 Jameson Taillon. In 2008, #3 Eric Hosmer is better than both #1 Tim Beckham and #2 Pedro Alvarez. #5 Buster Posey is better than all of them. But when you have years like 2009 with Strasburg, 2010 with Harper, or 2012 with both Correa and Buxton, their success will skew the data.
But in any given year, it's not really true that the #1 pick is clearly more likely to succeed than the #2 pick, or that #2 is more likely than #3, and so on. And even if it were true on a mostly yearly basis, you still have the issue of not having any way to determine where those players should have been drafted. Some people said Houston reached by taking Correa in 2012. So does he count toward #1 picks, or should he count toward later picks, since that's where he was projected to go? We will have the #3 pick, so if that correlation is actually significant, that should mean that no matter who we take at #3, he has a better chance to succeed simply because we took him at #3, no matter how good he is or where everyone thinks he should have been taken. So in that case, you still end up at not caring whether or not we 'reach' because the player has the same likelihood as anyone else we might take of succeeding because we would have taken any of them at #3.
And I apologize. No one should read this much about my thoughts on why 'reaching' a little doesn't matter in baseball.