So doesn't that sort of go against advanced statistics theory? Seems like a lot of people believe that actual shot quantity is what's most important because teams can't really mitigate for shot quality. I've read a few times "there's no evidence to suggest that any defensemen are better than others at preventing top scoring chances against". A poster on the Rangers board showed the team save % with Rangers' D-men and it was quite random (i.e. best D-men didn't necessarily have the best save % with them on the ice). So I would think that this works the other way too, no? It should be random on offense too, since the D-men don't really affect the shot quality against. Or is this an example of stats being misleading and advanced stats proponents taking them too literally?
Admittedly, I hadn't thought a lot about this and was just giving my impressions as a longtime DRW fan. However, I suspect that there is sufficient variation between the playing styles of individual players and individual teams to obviate a consistent relationship. Some defenders will block your shot (and some teams will make that the game plan for the whole team), while others will steer you to an area where your shot still gets on net but is relatively harmless. Some forwards are volume shooters while others only pull the trigger when they have an open cage. Looking at the same player or players over time may well yield more consistent results, but I'm skeptical about more general rules.
For the 95-96 Red Wings specifically, that was the year the Russian Five played together as a unit regularly. That group was more likely to make the extra pass looking for the perfect play than just fling it on net, to put it mildly.
I don't think the best defensemen necessarily will have the best on-ice sv%. Even as certain styles of defense would seem to make it more likely, there's a lot of other variation at play. But I don't think there's much question that defensemen do often directly impact the quality of the chance.