You can't access the science of others unless you're part of the game. It is only the molecular biologist who is publishing his own papers, getting invited to the conferences, having the discrete conversations with other fellow molecular biologists, who can capture the work of others.
The problems of access to journals are real. My PhD supervisor and other researchers I knew were making a real effort to work around them by publishing a lot of papers on public online "e-print journals" like arxiv.org
The problems of access to conferences on the other hand, I never saw. If fact, at every conference I went to while I was doing my PhD, there was always a handful of uninvited people who tagged along. Often they were nuts who believed they were close to a proof of something actually improvable and wanted to find a professor to show their work to. Occasionally they were just highly motivated amateur mathematicians. Either way, they were always at least tolerated from what I saw.
I've had the same experience at CS conferences, as one of the tolerated amateurs myself. Most of my idols in the field turned out to be pretty approachable.
The problems of access to journals are real. My PhD supervisor and other researchers I knew were making a real effort to work around them by publishing a lot of papers on public online "e-print journals" like arxiv.org
The problems of access to conferences on the other hand, I never saw. If fact, at every conference I went to while I was doing my PhD, there was always a handful of uninvited people who tagged along. Often they were nuts who believed they were close to a proof of something actually improvable and wanted to find a professor to show their work to. Occasionally they were just highly motivated amateur mathematicians. Either way, they were always at least tolerated from what I saw.