People want to publish in high-impact journals. Aside from name recognition (Nature, Science, New England Journal, etc), impact factor and similar metrics drive where people want to publish.
If you really want to strike at a particular journal or family of journals, you could work to convince academics not to cite articles in those journals. Since all these metrics are some variant of (inward citations)/(publications), usually over a 2-5 year window, this would have a tremendous effect.
I don't see this as a workable solution. You can't neglect to cite relevant research ideas simply because they were published in a closed journal, and there often isn't an open alternative to cite. I think you really need to have authors stop submitting to closed journals in the first place.
I'm fine with that characterization, too. I'm just saying that any barrier to access will shift the degree to which articles get cited away from what they (perhaps) "should" be based on their scientific value.
This is a really interesting point. I've definitely passed on reading papers that seemed potentially relevant because my institution didn't have a subscription to the journal.
Perhaps this is field-dependant, but I have never failed to find the full text of a recent paper I'm interested in even if the original paper is closed access - the paper is inevitably available on the webpage of the author, or their university/research group, or arxiv, or something like that.
For classic papers from the 1980s I have to bother with getting the VPN required for my university subscription access to work properly which is a pain, but I don't recall having any issues with publications from this millenium.
If you really want to strike at a particular journal or family of journals, you could work to convince academics not to cite articles in those journals. Since all these metrics are some variant of (inward citations)/(publications), usually over a 2-5 year window, this would have a tremendous effect.