Now, my argument isn't, “Stamkos is the reason that they lost”. My argument is that high-end goal scoring isn't the best indicator of winning games, and if your team commits a substantial % of the salary cap to house that high-end scorer, you could be caught with your eye on the business and not on the big trophy. When you work with a limited budget, as all teams are, your largest investments had better be ones that translate to post-season success.
The saber-metrics guys have come into hockey, and they have brought a suitcase of useful metrics for determining a players overall contribution to winning. Mainly, scoring chance +/- and Corsi +/- have revolutionized how teams look at which plays influence a game the most. The fact is, a players traditional +/- isn't an adequate way to evaluate contributions. This is because the league average goals/game is only around ~2.5; goals just don't happen enough in NHL games to be a good barometer. For example, Corsi +/- addresses this issue by looking at all of the types of plays that correlate to goals rather than goals itself. Teams will log the differential between shots from the slot, blocked shots from the slot, missed shots from the slot, etc. Teams sometimes adjust their equation for quality of opponents on the ice during each shift.
I'm just scratching the surface of the new metrics being used. Teams look at everything now; they'll log what spot the the puck was when a player gets on the ice for a shift, and where the puck was when he get off the ice. If that player tends to get on the ice in the defensive zone and quickly transitions it into the offensive zone, that player is assigned credited for doing so. If a player gets on the ice with bad teammates, and still transitions the puck, some teams will assigned credit proportional to how bad his teammates were.
Now, you asked me which players deserve to be paid the most, so I'll answer it: Pay the players that have regularly, and predictably made the greatest contributions to winning hockey games in both the regular season and the post-season. Understand that total points and age is not the full story on a players relative value.
Obviously, these arguments always take place in a vacuum, as scouting and savy contracts could theoretically negate/compensate for bad contract. However, if I was in charge and building a team, I would commit money to grittier players, two-way players, especially at the forward position. I think the most valuable players are usually in the conversation for the Selke award, guys like Datsyuk, Kesler, Toews. They all have appeared in the top 20 players in the regular season in on-ice Corsi, and the top 10 in the playoffs. They get better when the stakes matter.
At minimum, your highest paid player should have a net positive impact when he gets on the ice. But the fact is, Stamkos just doesn't. Yes, he scores a lot of goals, but his team doesn't get better when he is on the ice.
As far as pure goal scorers are concerned, only Ovechkin deserves what he's getting. Personally, if I were Tampa bay, I would consider moving him to wing.
I lost some steam there at the end there, but I have a hockey game tonight and I need to stretch out. End rant.
(interesting Q&A) http://www.puckprospectus.com/article.php?articleid=834