I think sometimes people put a bit too much stock in the cost of a win. Sure, it's a decent barometer, but there are other things to consider...
If you're paying big money for a Kershaw, Hernandez or a Lester, I think a good chunk of that value is going to typically be realized through postseason success. (While a rotation of say McCarthy/Latos/Buchholz/Kelly/RDLR might get us into the postseason thanks to our offense, how do you feel about that rotation going up against Price/Sanchez/Verlander, Kershaw/Greinke/Ryu, etc...?)
What's the value of being able to throw a Jon Lester out there in game 1, or a game 7? The value of starting pitching is certainly magnified in the postseason in comparison to one offensive player (unless that player is David Ortiz, of course).
And I think if you're the Boston Red Sox (or any team spending well into nine figures per year), you need to build your team under the assumption it's going to make the playoffs.
As an aside on WAR - while I think it is overall a good indicator of team success, it should be noted that Giancarlo Stanton's monster 2014 was worth less wins than Jonathan Lucroy's 2014. It was about a win better than Shane Victorino's 2013 (which according to WAR was worth about $35,000,000...)
Jon Lester's 2014 was worth around $30,000,000. Jonny Cueto's season was worth around $42,000,000. Kershaw? About $55,000,000.
Um, yeah...