Someone with that level of intelligence and ambition today would have some interesting challenges, I think. Published journals are expensive and exclusive and inconvenient to get, so maybe he'd ignore those entirely. The internet, no doubt, has a firehose of such info from many sources worldwide. You couldn't consume more than a percent or two of the stuff being published on any given day. And you'd still want to tear yourself away and spend some uninterrupted hours on your own work.
In some ways, it seems like Newton had it pretty easy!
The doom-and-gloom about modern times is unnecessary. Where there is a will, intelligent minds will find a way. Hell... a motivated high-schooler can moonlight in a grad school lab and (pretty much) any of us would dole out authentication credentials for him to download whatever papers he'd like...
What I'd like is "published papers filtered for importance" (perhaps in a citation-count pagerank sense) combined with access to the actual content of the papers.
RSS or Atom. PDF or HTML. Ability to set thresholds would be nice too.