We need journals because without them we'd have the Android App Store. I know that seems like a stretch but the challenge is that there are lots and lots of papers (publish-or-die is a mantra remember) and so quantity is assured, but quality is more nuanced. And more importantly quality of the work. So what a journal does is builds a reputation for filtering for only very high quality papers, and then other systems feed off that. So if you publish in Science that is cooler or more impressive than publishing in a journal which has no standards, or worse your own self published journal. The Android App store has a similar situation except it doesn't (yet) have a good quality filter, there is no third party that has risen to prominence as the place to go for authoritative reviews or opinions on the quality of a particular application.