Personally, I'm not a fan of hypotheticals like that because I feel that people tend to put too much emphasis on the number of championships that players win in team sports in determining greatness. Now, I'm not suggesting that no importance should be attached to contributing to contributing to those championships, but it shouldn't be the be all end all that it is for some. I once had a friend get fighting mad at me because I wouldn't agree with him that Henri Richard was the greatest center in Habs history, simply because of the 11 Cups. Pointing out the greater personal impact of players like Beliveau and Morenz, or even that Beliveau won the Cup 10 times made no impact, because he was so blinded by that number.
Now, I'm not making that accusation in this particular case; I only bring it up to demonstrate the general tendency. But, that tendency's existence makes me respond with this question: If Roy would be in a top five, why isn't Beliveau now? Okay, I know he doesn't have the number of Conn Smythes, but he's generally considered to be the obvious winner in 1956, before officially winning it in 1965, not to mention the fact that he captained half of the ten champions he played for. On balance, I think that's fairly comparable to the theoretical Roy resume here. And as for ignoring the domino effect, can we do that? There are too many possibilities to allow that. How many other players have their number of Cups run up? That's dependent partially on whether the trade to Colorado still happens. How does it affect players on teams Roy doesn't play for? Does Lemieux end up not winning a Cup because Roy's Habs took it in 1991 and 1992? Man, that's messy.