I agree with your thinking, but another way to acquire talent is taking a chance on a promising youngster.
Kulak is more valuable then a high second round pick. Hes young, appears to have offensive potential, playing a good amount of minutes, and has arguably been the most solid d man for the past 4 or 5 games (not a high bar unfortunately). Throw in the fact that Calgary (in your scenario) shows interest in keeping him, and willing to take a hit to do so.
Id way rather lose Ferland than Kulak, if Kulak keeps playing like this.
Can't say I disagree that on paper, it's much more logical for the Vegas Casino Robbers to just take the best talent available. The only thing that sticks out though is that historically, the prices to protect players has been
extremely low.
In return for agreeing not to select certain unprotected players, the Blue Jackets and Wild were granted concessions by other franchises. The trades not involving Blue Jacket or Wild draft picks were booked as being for "future considerations":
Columbus
San Jose traded Jan Caloun, a ninth-round pick (Martin Paroulek) in the 2000 NHL Entry Draft, and a conditional pick (Aaron Johnson) in the 2001 NHL Entry Draft to Columbus on June 11, 2000, after the Blue Jackets agreed not to select Evgeni Nabokov.
Buffalo traded Jean-Luc Grand-Pierre, Matt Davidson, and two fifth-round draft picks, one each in the 2000 (Tyler Kolarik) and 2001 (Andreas Jämtin) Entry Drafts, to Columbus on June 23, 2000, after the Blue Jackets agreed not to select Dominik Hasek or Martin Biron.
Minnesota
San Jose traded Andy Sutton, a seventh-round pick (Peter Bartoš) in the 2000 Entry Draft and a third-round pick (later traded to Columbus - (Aaron Johnson)) in the 2001 Entry Draft to Minnesota on June 11, 2000, for an eighth-round pick in the 2000 Entry Draft after the Wild agreed not to select Evgeni Nabokov.
And given how the prospect pool is structured, the Flames can more easily afford to lose a player like Kulak (Kylington, Andersson, Hickey, Fox, Culkin) than someone like Ferland (Smith and Carroll).