>Most academics tend to self-host preprints, or stick them on arXiv so the publisher as a hosting service isn't really the "value-add" offered.

That is not true for chemistry or biology. Tell them what arXiv is and they will say, huh?