Reading back over this thread (after it was bumped) taught me a lot about Hasek's achievements pre-NHL, which I knew little about. Very interesting.
I have no dog in this fight -- I'm generally kind of dispassionate about goaltenders -- but I'll just add my thoughts:
-- I really don't think analyzing goaltenders' save-percentages (or any such stat) in the 80s to mid-90s is a useful exercise. After that era, when we get into Hasek's late-prime, it becomes more useful because the NHL teams were getting more stream-lined, there was more parity, and every team started playing the same style; i.e., all teams became more comparable. But c. 1985 to 1995, I don't really trust save-percentage or GSAA and whatnot as truly indicative of anything BECAUSE EACH TEAM's STYLES WERE SO DIFFERENT. Patrick Roy's numbers were favored by his joining the best defensive team in the 80s (Montreal) when he entered the NHL. Hasek's were not favored by joining Buffalo, which had middling-to-poor team defense, but later in the decade became defensively better as they built a team around Hasek. Conversely, when Montreal sagged a bit in the early/mid-90s, and after Roy joined high-octane Colorado, his numbers and statistical dominance also dipped a lot, as you would expect.
But raw stats simply don't tell the whole story anyway. The small difference in numbers is fairly trivial, I think, between any remotely comparable goaltenders of that era. Some teams regularly had 3-goal leads in the third period and let-up (leading to more goals against, i.e., the Grant Fuhr type), and some teams were so bad they had lost the game by the second period and subsequent goals against were meaningless (i.e. the Leafs in the 80s, Quebec early 90s).
The only time the stats become quite meaningful, I think, is when there is an extreme outlier - for example, Roy in the 1988-89 season, or Hasek's best save-percentage for (I think) six seasons in a row. But I really don't think there is any point in doing in-depth analysis of goalie-stats from that era (or indeed, most eras), if the numbers are even remotely close.
-- Patrick Roy's NHL playoff-performances were incredible and can't be under-estimated. I suspect Hasek's would have been similar (they aren't that bad, as it stands, with two Cups), but for the teams he played on -- the hand he was dealt, if you will. (I'm sorry, but I refute any argument that Roy's teams weren't way better in player-personnel than Hasek's.) However, the facts are the facts, and Roy was the superior playoff goaltender, with the numbers and Cups to prove it. On the other hand, Hasek came up in the 70s and became a big-time player in the 80s in the Czechoslovakia, where NHL success was a pipe-dream and the Olympics and World Championships were the #1 priorities. It perhaps isn't shocking that Hasek's Olympic/World Championship results are better then Roy's, while Roy's playoff results are better than Hasek's.
-- The one time they faced each other while each was in his late-prime (1998 Olympics), they both were superb, but Hasek won.
In short, I think it is very easy to make a strong case that Hasek is the greatest goaltender ever, from any of the perspectives of Olympics, world championships, NHL regular season, or NHL playoffs. Only on the last of those 4 does he fall behind Roy a bit.
It's one of those things, though -- Hasek led the '98 Sabres to a great season and the '99 Sabres to game 6 of the Finals. What else could he possibly have done other than scoring a hat-trick himself? And once he did get to play on a high-octane team, he immediately won the Cup.
I personally think Sawchuk, Roy, and Hasek are the three best claims for greatest goaltender. (I'm not personally bothered who is 1, 2, or 3.)