It comes down to Roy was beating out mediocre talent for his Vezina trophies in an era that is basically a desert for good goaltenders outside of Roy, while Hasek was beating out guys *like* Roy, but also Brodeur, Belfour, and CuJo.
One of those things that came up recently:
Ranked Save Percentage Finishes (Before 1994)
Belfour: 1, 3
Hasek: 7
Joseph: 1, 2, 6
Roy: 1, 1, 1, 1, 2, 5, 8
Ranked Save Percentage Finishes (1994-1999)
Belfour: 8, 10
Brodeur: 3, 4, 5, 7
Hasek: 1, 1, 1, 1, 1, 1
Joseph: 6
Roy: 3, 4, 7, 8, 10
Belfour, Joseph, and Roy took every save percentage title from 1988-1993, but between the three of them, they had just two top-5 finishes from 1994-1999, so it isn't as though Hasek was beating these HOF-level players when they were having their HOF-level seasons.
They were still marquee names from 1994-1999 largely due to playoff performances in that era and the reputation they earned in the early-1990s, but they didn't necessarily make for consistently excellent Vezina competition.
Hasek's top competition for his save percentage titles were John Vanbiesbrouck, Chris Osgood, Daren Puppa (a title Hasek earned while sitting out with an injury while Puppa's numbers dipped during a restless playoff push), Jeff Hackett, Tom Barrasso and Trevor Kidd, and Byron Dafoe.
He wasn't beating the best of Belfour, Joseph, and Roy from 1994-1999 any more than Roy was beating the best of Hasek and Brodeur in 1991-92; they just happened to be in the same league.
That Patrick Roy - for stretches - could play at a similar statistical level to Hasek after his 1988-1992 peak (notably the 1993-1997 and 2000-2001 playoffs, much of 1997-98, 2001-02 and the final half of 2002-03) suggests that he was
capable of playing like the best goaltender in the world, but couldn't
consistently do so in the same manner he could from 1988-1992. And I think the only alternative to this theory would be to assert that Patrick Roy was a better goaltender in the final three months of his career than he was in 1988-89 when he went 25-0-4 in the Montreal Forum.
Hasek's statistical record is largely protected by a focus on averaging statistics that ignore a consistent streak of injury - first in 1992-93 when he became a starting goaltender, necessitating a trade for Grant Fuhr; next in the 1997 playoffs, leaving Buffalo's fate in Steve Shields' hands for both playoff rounds; again in 1998-99, missing over a month at the end of the season and parts of the Conference Finals; a multiple-month stretch in 1999-00 that gave Martin Biron the reins; a forfeited 2003-04 that Yzerman described as "very strange" following Hasek's 2002-03 retirement; and finally a 2005-06 season that saw his Senators teammates and coaching staff beg for him to play.
More than half of Hasek's eleven seasons where he finished top-10 in save percentage (second most in the official era behind Patrick Roy's 15) resulted in either a month-long+ injury in the regular season (the sort of thing that would knock a skater out of a top-10 scoring race) or missing his team's playoff run. It's the sort of thing for which great skaters like Malkin and Forsberg get dragged over the coals, but since save percentage is an averaging statistic, Hasek largely receives a free pass.
Hasek could stop a puck like nobody's business, but there's a reason the same media that consistently voted him the best goaltender in individual seasons also consistently rates Patrick Roy or Terry Sawchuk or Martin Brodeur as better all-time.
If given the choice of a 6x 1st Team All-Star (1994, 1995, 1997, 1998, 1999, 2001) with 5x top-3 Hart balloting or a 4x 1st Team All-Star (1989, 1990, 1992, 2002) and 2x 2nd Team All-Star (1988, 1991) with 5x top-5 Hart balloting, they both are amazing, but the first guy sure sounds better in those six years.
...but when the second guy also has two additional seasons with Vezina nominations (1994, 1997), and three additional seasons with Conn Smythe Trophies (1986, 1993, 2001), you end up with that second player getting invited to twice as many NHL Awards ceremonies. Patrick Roy won trophies 8 years before Hasek's first NHL nomination, and he won trophies the year after Hasek's final NHL nomination. And I don't know that any 1980s/1990s what-ifs about a goaltender that couldn't capitalize on opportunities for a starting job in 1990-91 (lost to rookie Belfour), 1991-92 (during Belfour's holdout), or 1992-93 (another injury) bridge that kind of gap.