More importantly the CUP notes that their website tracks "millions" of hits to their Abstract pages which turn away otherwise empty handed.
I'm no genius but the iTunes experience would seem to provide a simple path for publishers. Drop the price point (for full access, mind you) until the market responds. Maybe that is £3.99, maybe £0.99 or maybe even less. I don't know but there is very likely a nominal rate that gets those millions who are currently turning away to pay for the article.
Or maybe the Netflix model would work better. Again, the cost is going to have to be reasonable. I read a lot of Elsevier content but still the barrier has to be low. £50.00 per year for (real) access to *every* journal? I might even do that just to cover my browsing when I don't want to wrangle with VPN and proxy servers.
One of the interesting things to arise in this recent round of OpenAccess discussion, in my mind anyway, is the role OD science blogging. Especially the Researchblogging.org style which focuses on explicating, you guessed it, research articles. What great advertising for publishers! Free product shilling from a small but generally dedicated class of folks.
Even Ed Yong may not be able to write purty enough to get the casual reader to part with £3.99. But to part with £0.25? Maybe that would be possible.