To just elaborate on the preferential ballot thing, there is of course no way to know whether or not the old system would have produced similar winners, but it's worth examining how the recent splits might have something to do with the new way Oscar votes for Best Picture.
I'm going to say that the Argo/Life of Pi split was just a fluky situation all around, and Argo would have won regardless on a simple, vote-for-your-favorite ballot. Had Affleck been nominated, he most surely would have won Director, and Argo was the favorite all season. I think this outcome was due to a quirk from the Directors' branch.
However, the other two splits -- 12 Years a Slave/Gravity and Spotlight/The Revenant seem like exactly the kind of scenarios you could encounter with this system. Just look at this board as a sample pool: Gravity had fierce partisans but also some pretty strong detractors; a much wider consensus of posters here were happy to accept 12 Years as a Best Picture winner. I don't know that either Spotlight or The Revenant provoked similar enthusiasm around here, but many of us were dreading a win for The Revenant; Spotlight, in contrast, was a movie a lot of us felt was a perfectly solid choice, even if it wasn't our choice.
It's still interesting though, how often we might have had splits in this new era, but didn't -- Avatar/Katheryn Bigelow, The King's Speech/David Fincher, The Artist/Martin Scorsese, Boyhood/Iñárritu, Birdman/Linklater all strike me as outcomes we could have had, but didn't. Oddly, the only Best Picture winner in the more-than-five era that feels like it definitely COULDN'T have had a split was one that did by force, and that's Argo, in a race that would have seemed like a total runaway had Affleck gotten his expected Director nod.
Forgive me if someone else has already mentioned this here -- a lot of Oscar coverage over the past day has bled together in my mind -- but it's also worth noting how the trend in number of Oscars for Best Pictures has decisively gone down in recent years. Spotlight pulled it off with a paltry 2; 12 Years a Slave and Argo only managed 3; The King's Speech and Birdman, in this context, looked amazingly healthy with 4; and The Artist was a virtual sweeper with 5. Compare that to the '90s, when three winners scored 7, and others reached up to 9 and 11. It seems that there has been a real change in terms of how much across-the-board support a movie needs to win Best Picture.
Part of me wonders if this has to do with different kinds of movies winning Best Picture. And there's definitely some of that -- Spotlight resembles the kind of movie that was often an also-ran in the '90's, like The Insider or Quiz Show, rather than the winners. But others seem like movies that could have swept up more Oscars, but just didn't. Imagine if Gravity hadn't existed in 2013 -- 12 Years a Slave certainly would have added a well-deserved Director prize to its haul, but it very likely could have won Cinematography and Score (categories it didn't even get nominated in!), to say nothing of both male actors and the design categories being within reach. And many of us predicted The King's Speech for Score, Production Design, AND Costume Design, which would have given it a haul much closer to Shakespeare in Love's than Rain Man's. Or, to flip the argument, perhaps some of the movies that have dominated down-ballot in recent years (like Hugo, Gravity, and Mad Max) are just a bit more fantastical than the historical epics that were routinely sweeping in the '80's and '90's -- had one of these films been even slightly more serious, might it have gone all the way to Best Picture?