Part of the issue is that you're assuming Tkachuk pans out. It's still early, I mean, 2 games... remember when Lee was going a top 4 dman and Cowen a partner for Karlsson?
I don't think that's a good comparable to be honest, I'd perhaps compare an average draft to a strong draft; for example, would you trade one of Huberdeau (3rd) Larsson (4th) or Zibanejad (5th) in the 2011 draft for an unknown top 10 pick in the 2015 draft: Could be any of McDavid, Eichel, Strome (the not good one), Marner, Hannifan, Zacha, Provorov, Werinski, Meier, or Rantanen? There's a chance you downgrade, but the drop off isn't unmanageable, you still end up with a great player (unless you get Strome), but you also have a chance at some elite guys that are definitely (well, likely, I can dream that Tkachuk surpasses Eichel and McDavid... right?) better. In that scenario, an known choice in 2011 vs an unknown field in 2015's draft, I could be easily persuaded to take the risk (in this case, with the benefit of hindsight I guess).
The thing is, while the stats you quote are averages, we know that the next draft is much deeper at the top end, and this draft had a big drop after the top 2. I happen to be one of those who defended Tkachuk as an option before we picked him, but I could see the rationale behind preferring keeping the 2019 pick back then, and it's made even more appealing now that we know Hoffman and Karlsson are gone.
Anyways, I don't think it's a no brainer, but you don't get many chances at top 2 or 3 talent, so imo it's worth the risk. Now, the team may see Tkachuk as top 3 talent, so there is that too.