Drawing the line between different types of access
In response to your editorial Communicating science in an electronic era, I suggest prominently and explicitly distinguishing open access no-toll access from either lower-toll access or access-toll subsidy.
And we clearly should distinguish also the research access problem – which is a primarily a problem of the research community — from the journal cost or journal budget problem, which (despite the weak links between the two problems) is primarily a problem of the library community and the publishing community. Researchers have an interest in the latter, to be sure, but lower-toll access is not only not a solution to the problem of open access, but, as oft-noted, a trojan horse, likely to wed us more strongly to the toll-based status quo.
The bottom line to always remember that even if all journals were non-profit and sold 'at-cost', the toll-barriers and access-problems would still be there, and it would still be true that all research was inaccessible to most of its would-be users — a circumstance that is no longer either necessary or tolerable in the online age.