I like Skinner too. He has the potential to be in the NHL as soon as next season and fantasy-wise could very well be an impact player in 2-3 seasons. He has the skill to pot 40 goals in the NHL one day.
I was surprised but quite satisfied when I managed to snag him with the 20th overall pick in my 30-team keeper league entry draft.
I tend to not draft defensemen during entry drafts (for fantasy obviously) because they tend to be much more of a question mark then forwards. The top few defensemen (Fowler, Gormley, Gudbranson, etc) are all going to be solid NHLers, that's for sure. But fantasy-wise, their production is far from assured - they could top out at 30 points or at 60 and still be effective NHL players either way. For that reason, generally avoid drafting NHL defensemen at the entry draft - I prefer to go after them 2-3 years after their draft date, because by that point it becomes much clearer what kind of player they are developing into and what kind of role they'll get on the team that drafted them.
The other issue with young defensemen is that they take far longer (in general) to develop then forwards do. The top drafted forwards tend to be top-line, valuable fantasy players 2-3 years after they're drafted; the second-tier draftees probably 4-5 years afterwards. With defensemen though, it's rare to see a player produce meaningful numbers before they hit 23. Sure, there are exceptions like Myers and Phaneuf, but these are few and far between. Guys like Karlsson and Carlson could produce early too, but again, these don't occur very often. More often then not, young defensemen at the NHL level work on their defensive skills first so as not to be a liability on the ice, and then, after establishing themselves and earning confidence in themselves and the coach, they may begin to take offensive risks and produce more. Whereas with forwards, they are generally given the green light right away to take offensive risks (centres are generally the exceptions as they have more responsibilities to cover), and therefore will start producing more, earlier.
If you're going to draft a defenseman, only take one that is clearly an offensive defenseman. Fowler, for example, is definitely an offensive defenseman, though he does have some two-way qualities as well. Other examples of offensive defensemen in this year's draft include Joey Laleggia (also drafted by me), Stephan Silas and Ryan O'Connor. Guys like Gormley, Gudbranson, Pysyk and McIlrath are far more two-way and will most definitely be making the NHL in a more defensive role first. Eventually, they will become effective, point-producing NHL defensemen, but that process takes time - often 3-4 years. We're seeing that with guys like Hedman, Erik Johnson, Kulikov, etc.
So back to the original topic... with that #3 pick I would probably take Skinner or Fowler, but Granlund, Tarasenko or perhaps Connolly would all be reasonable as well. Tarasenko will probably be the best forward available long-term behind Hall and Seguin, but he'll be in the KHL for the next 1-2 years and perhaps longer. Connolly is a bit of a risk with his injury and all, but he'll probably pan out to be a pretty effective offensive force. Granlund is the safest pick IMO, but I don't think he possesses quite the same upside that Skinner, Tarasenko or Connolly do.