A good way to deal with the issue is for reputable academics in a field to start new journals, and do all the work they did for the old paid ones for free, but

A good way to deal with the issue is for reputable academics in a field to start new journals, and do all the work they did for the old paid ones for free, but this time for a free end-product. This exists: I know of one that's been going a decade or more, hosted and started at USC, and there are probably a lot more. Even better, to support this, have an open-source freeware Journal CMS and publishing system for academics to use as a plug and play solution.