ArXiv has deliberately left the implementation of such systems to others. There's nothing stopping anyone from making such an overlay, should they choose to do so (and many have tried).
2 and 3 are self-reinforcing effects -- why does Nature attract prestigious papers? Because it publishes prestigious papers. Why does Nature attract high-quality research? Because it is full of high-quality research.
All of this is to say that it is likely that, given a distribution of reasonably-edited arXiv-overlay journals, there's a pretty good chance that just a few of them in any given field (or at large) will attract most of the serious work