Journals are mainly interested in profit, not fixing anything.

That should be only sort of true.

Some of the very high-profile journals are run by non-profits, including: Science (American Association for the Advancement of Science), PNAS, (National Academy of Sciences), eLife (HHMI/Max Planck/Wellcome Trust). A slew of more specialized journals are run by societies too.

In theory, they should be willing to lead the charge. In practice, I think they are largely dependent on income from the journals for a lot of their operations and so are reluctant to rock the boat.

Surely publishing a result is not in itself costly. But I guess the peer review is.

So journals could have a section (the grey pages?) for "unsellable results" that they didnt give a peer review. They would of course need to assess them in some other way, to ensure a reasonable level of quality.