There's no reason to not count both overall shot attempts against and scoring chances against. Some teams, like LA, focus on limiting time and space inside the defensive zone to force turnovers. Just counting scoring chances wouldn't accurately represent the defensive impact of their players.
Problem with looking at relative sv% is that it's a hugely volatile metric, driven primarily by uncontrollable factors. Including it is more likely to infuse misinformation than actually help the formula. As an example, Bergeron has a season where he had one of the worst relative sv% in the league, the worst on his own team. Besides, it shouldn't be needed because what it represents will be covered by looking at high quality chances.
I'd also skip the successful transition. It's a slippery slope to start to include transition elements in a measure of defense. That's one of the reasons I don't like using possession based metrics as proof of defense. It's more a measurement to how little they have to play defense, rather than how good they are at it, in my opinion.
Forced turnovers, percentage of puck battles won, number of won battles and puck recoveries relative to team and possession could be useful. Some guys are just vastly more involved than others. In Toronto we had Dion who was pretty successful at what he did, it's just that his skating meant that he never got to engage as often as you'd like.
You could also look at something like percentage of shot attempts blocked. It would mostly show how successful the player is at blocking the puck when an attempt is necessary.
