They've been plenty explored, discussed, chewed over, and more, including the scenarios your presented. The costs have been analyzed in dozens of different ways, with many business models proposed and some executed upon.
(Note that your mathematician gets indirect funding by having access to the university library. As a non-academic, I can use the local college library but must pay access fees for some services that are free to staff and students. At a somewhat further away university library, as a visitor I can read journals online but am not permitted to make copies.)
Nor is our "current system", concentrated as it is in the hands of Elsevier (and its 37% profit on revenue), all that old. Most people outside the big publishing companies didn't fully explore the downsides of moving away from the system we had before the 1980s - or at the least, nothing like the ongoing discussions concerning open access.
(Note that your mathematician gets indirect funding by having access to the university library. As a non-academic, I can use the local college library but must pay access fees for some services that are free to staff and students. At a somewhat further away university library, as a visitor I can read journals online but am not permitted to make copies.)
Nor is our "current system", concentrated as it is in the hands of Elsevier (and its 37% profit on revenue), all that old. Most people outside the big publishing companies didn't fully explore the downsides of moving away from the system we had before the 1980s - or at the least, nothing like the ongoing discussions concerning open access.