> My field (computer science) and a few others at least don't have the problem of paywalls. Authors always make their works available; a few publishers have relaxed their copyright policies and others have an implicit promise not to sue. It's not ideal, but it's not too bad.

This depends very much on the age of the paper. The older it is, the less likely that it is on the web. If anything it reinforces the profound cyclical amnesia of our industry (never mind academia).