This isn't necessarily true. At a certain point, wins for a baseball team become more important (somewhere around 88-94 wins), because they're harder to get and they're more likely to get a team into the playoffs. In light of that, acquiring a player who adds a 2 wins -- say, 6 WAR vs 4 WAR -- is worth more money when you have an 87 win club. It all gets a bit complicated, but at a certain point the extra talent is worth the additional dollars. This is part of the reason people believe the asset cost was worth it for the Jays when they acquired RA Dickey: it may have pushed them from a team around 86 wins to about 89-90 wins.
It's difficult to adapt that reasoning from baseball to hockey for a number of reasons:
1) Obviously baseball doesn't have a cap. Teams can value wins at a higher amount because they aren't constrained by limitations on their spending other than money. The change in the win curve happens because teams can generate a ROI on their investment above and beyond what most teams can. With revenue sharing, I don't think NHL teams make nearly as much from a single playoff series.
2) Baseball has a lot more variability in the playoffs - pretty much any team that makes the playoffs can win it all with almost equal likelihood as a seven game sample in baseball tells you a lot less about true talent than it does in hockey - but half the number of teams actually make it there, so getting to the playoffs is significantly more valuable than it is in hockey. This is what creates the additional value depending on where you are on the win curve. In hockey, you don't have that same impetus. Teams are not constrained by playoff spots, as it's incredibly unlikley that a team facing the prospect of pushing themselves over the win threshold for the playoffs is going to be competitive for the Stanley Cup. Just as in baseball, a 95-100 win team doesn't gain a lot of value from adding additional wins, teams like the Canucks, Kings, Blackhawks, Blues, etc. don't either. The win curve idea would apply to teams like the Flames and Stars that are always on the cusp of the playoffs.
3) Baseball essentially has one game state while hockey has three. You can maximize value in hockey by getting players that specialize in different areas (EV, PP, PK) to maximize your competency in all three without having an elite player in any one area. In baseball, a player is going to be either doing one thing (pitching, DHing) or has to do both (play defence and hit). In hockey this would be like making you play defencemen in equal amounts (proportionately) on the PP, PK, and EV. Or another way to equalize things would be to allow a significantly expanded roster in baseball so that teams could constantly change players to take advantage of platoons and defensive/pinch hit substitutions.
As for committee scoring, I guess I'd say that if you could upgrade a 5g/40pt defenseman to a 15g/60pt defenseman, you're adding a lot of surplus value into a single roster spot. That's worth something.
This is the big value in elite players, IMO. Consolidation in value is much more important for good teams than bad, and, with the NHL cap structure, elite players tend to get undervalued in the market moreso than average to above average players. Unlike in baseball, you don't see an (essentially) linear progression in salary depending on the number of wins you add. Instead, top end players are constrained by the percentage of the cap they can take up and I think this acts as ceiling that no one wants to hit.
This means that there is relative value in the $7-8 million true superstars compared to the $6-7 million very good players and the $4-5 million above average players.