Still beats picking a less talented and less intelligent player that won't make the league at all.
I prefer drafting players with some size in rounds 1-2 but throwing darts at smaller players with NHL skill and IQ in rounds 3-7 is smart practice when looking at the history of players drafted late that ended up being above replacement level (top 6 forward/top 4 D).
One of the challenges is there is only a finite number of roster spots. Once a team has a core established and locked up to term, it impacts their drafting strategy. Each prospect needs to be evaluated as to where they could fit in that future lineup. If they are a high skill prospect then the question is whether they have the kind of upside to displace or make expandable one of the core pieces, and if they don't or are unlikely to do so, then the question is how effective can they be in a lesser role.
Once that core is established then it is more important to build out the right support cast for that core. That means prospects that can play effectively in more roles are typically more valuable than players that can play in limited roles. That is unless a prospect looks like they will be elite for a very specific role and sometimes a very limited one. This would include a lot of depth roles like penalty killers, agitators, 3rd and 4th line centers, etc.
The implication of this is that a skilled prospect who likely won't be able to challenge the core to displace their roles and is more limited in their ability to play multiple smaller roles effectively, is not as valuable of a prospect as those that are more capable of playing support roles effectively.
The common counterargument to this is the suggestion of drafting skill and then trading for or signing support players. There are two problems with this counterargument. The first is that most skilled players that aren't able to displace core players and can't play support roles very effectively are not that valuable on the trade market. All teams would like to win the Stanley Cup and are looking to build their rosters to achieve that objective and as a result they don't want to pay premium assets or even very valuable assets for a player that won't be part of their core, will likely get pushed down the lineup and won't be as effective as other players in a support role.
The second is that high end support players are difficult to acquire while they are in their prime and while they are cost controlled. Particularly in a salary cap era, there is an importance on controlling and optimizing the cost of the support role players to build around the more expensive core. Drafting these support players allows an organization to have them contribute to their team while they are likely in their prime and while they are cost controlled. In addition, if they are high end support players or elite for specific roles then they are highly coveted by other teams and will garner a high return in a trade at the trade deadline if the organization is forced to go through a rebuild and is looking to sell off.
As a result, this impacts the draft strategy of most organizations once their core is pretty much in place. If they are good at this drafting approach then they stay highly competitive for multiple years and either win or have multiple attempts to win one or many Stanley Cups.
