I agree that the filter is where the most value is, but it’s wrong to assume that the publisher is providing it. They’re not. The science community is the ones

I agree that the filter is where the most value is, but it’s wrong to assume that the publisher is providing it. They’re not. The science community is the ones providing it — for free. Same with publishing templates. The science community itself provides the style guides and style templates. The publishers provide printing presses. That’s it. The scientific community can decide to abandon the publishers, and start competing journals, which will automatically transfer the heft away from the original and to the upstart, because it is the community that provides the value. This may sound idealistic, but this has already happened. The linguistics community abandoned Elsevier’s Lingua and replaced it overnight with Glossa. The linguistic community’s success, then inspired the mathematics community to abandon Springer’s Journal of Algebraic Combinatorics and replace it with Algebraic Combinatorics. This path is traveled. The pattern established. In 2019 it’s just much easier to start a publication since you need very little economic capital.