I don't see why it would be a controversial bill in Congress to simply exempt scientific publications (broadly construed) from all copyright law. Springer and Elsevier (the latter not even an American company) would be losers, but I doubt they spend more than a couple $mil on lobbying per year between the two of them. The benefit to humanity seems pretty clear.

This seems entirely within the scope of Congress's power under the Copyright Clause, and need not be tied up with other more controversial aspects of intellectual property law. There was a time when one could have made a case that permitting copyrights on scientific publications would "promote the Progress of Science and useful Arts" (e.g. because only journal publishers were in a position to efficiently provide copy editing, nice formatting, and distribution). None of that has been relevant for decades, so it seems to me that one could even make a case that the status quo violates the Copyright Clause, since enforcing journal copyrights pretty clearly diminishes rather than promotes the progress of science.