I disagree, I cite important work all the time and you definitely need the original paper. Paraphrasing isn't really done enough to negate the need for the orig

I disagree, I cite important work all the time and you definitely need the original paper. Paraphrasing isn't really done enough to negate the need for the original paper, otherwise what kind of lame plagiarized paper are you writing? Citations only tell you where to look, and if that article is gone then you're screwed. What do you mean by "duplicated elsewhere"? Sitting on some scientist's hard drive doesn't count, it has to be discoverable. That's what the issue is: when journals die, how do we ensure that the papers are saved somewhere easily searchable?