Goalies have to be judged on a case to case basis, and the only real way to evaluate them is by watching them. It would erroneous to simply state that the goaltender win the highest save percentage is the best, or that the goaltender with the most wins is the best, because neither statistic is specific enough to the goaltender itself, and are both prone to being influenced by outside factors. Unfortunately, it's just not that simple.
Perhaps in other sports like baseball, statistics can tell most if not all of the story about a player. Using wins, ERA, WHIP, opponents batting average, etc., one can make a fairly accurate statement about how good the pitcher is, as those statistics are very specific are for the most part only influenced by the pitcher itself.
With hockey, and specifically goaltenders, you can only take the numbers game so far. The main flaw with GAA is that a goaltender can only do so much. It would be unfair to suggest that even the best goaltenders in the game could have a great save percentage on the worst defensive team. Most would consider Roberto Luongo to be one of the top-5 goaltenders in the game, but his GAA was 2.43 in '03-'04; a far cry from the league leaders. Goaltenders are only human, they cannot do it all alone; playing behind a bad defense will hurt these statistics. Save percentage is better than GAA, but still flawed, because it's nothing more than a math problem masquerading as a hockey stat. A goaltender on a great team could stop 9 out of 10 shots and have a save percentage of .900, while another goaltender on a bad team could stop 36 of out of 40 shots and have the exact same save percentage; .900. If you chose to look purely at the save % number, you may be inclined to believe the goaltenders are on the same level, which in fact could be very misleading. If the goaltender on the good team is only facing 10 weak shots, yet the goaltender on the poor team is facing 40 strong shots and has the same save percentage, it would logically make sense than goalie B (on the poor team) is doing much, much better. On top of that, what if a goaltender only stops 25 of 30 shots (.833 save %) but all 30 shots are excellent scoring chances from great shooters. At face value, .833 is unimpressive, but it hardly tells the whole story. In the end, Wins is the only statistic that matters, but it isn't the better stat to judge goaltenders. I'll go back to Luongo for a second: He had 25 wins in '03-'04 while Dan Cloutier had 33. Kudos to Dan for winning on a good team, but I don't know of anyone, even the biggest Canuck homers, who would take Cloutier over Luongo. Thus, even wins is a flawed stat when it is used to evaluate goaltenders. In the regular season.*
However, this is not to say we should punish goaltenders who happen to be on great teams. There is something to be said for playing well on a good team during the regular season, because after all, there have been many who've failed at just that. To me, putting up great stats (save %, GAA) on a losing team doesn't mean any more to me than putting up bad stats (save %, GAA) on a good team while getting a lot of wins. They're both equally ambiguous. These statistics are here to give us a way to get a feel for how a goaltender is playing, but not to be the ultimate determining factor. Regular season win totals must be taken with a grain of salt as well, as winning a lot of games may not be enough. It must be accompanied by strong play. All of the major goalie stats have flaws, and the only real way of finding out who is playing better is to use your own two eyes.
*Note: Things are a bit different if we're talking about the playoffs, in regards to wins. Wins are by far more important than any other stat for post-season play. In the playoffs, any one team can beat another, it's a completely different ball-game. There are no nights off, nor are there easy opponents. When you win in the playoffs, all it means is that you did exactly what was required and got the job done. No more, no less. Perhaps you won 6-4, or perhaps you won 1-0, but if you got the W, you did your job. You might ask, why is this different than the regular season? Well, the playoffs are vastly different; every game is a battle, wins are much harder to come by, and the game are vastly more important. Once the playoffs roll around, what you did during the regular season doesn't even matter. Are wins the "be all and end all" in the playoffs? No, not at all, but there are much more important than any other stat.