The recent launch of several high-profile open-access journals by commercial publishers including Nature Publishing Group and SAGE elicited cheers from veterans of the open-access movement.
Here, they thought, was evidence that their ideal of making research freely available online, as expressed in 2002's landmark Budapest Open Access Initiative document, was finally gaining mainstream traction.
But according to Christopher Pressler, director of research library services at the University of London, the enthusiasm for "gold" (journal-based) open access is relatively recent and amounts to a "fundamental compromise" necessitated by the open-access movement's failure to plan for the financial sustainability and academic appeal of "green" open access, which is built around self-archiving papers in institutional repositories.
He contended that librarians' original goal in pushing for open access was to take academic publishing away from the commercial enterprises they perceived to be making unjustifiably large profits from charging universities for access to their own research.
Mr Pressler told Times Higher Education that the rise of gold open access could lead to an even greater drain on university budgets due to some publishers' alleged practice of "double dipping": charging authors for open-access options without reducing subscription prices proportionately.
Institutional repositories have been established in their hundreds by institutions worldwide, including around 150 in the UK, but by common consent have attracted disappointingly little content.
"Setting up repositories was one of the major strategic ambitions in the past 10 years, but we have to face the fact that they have not transformed the research landscape in the way we had hoped," Mr Pressler said.
But not everyone is ready to write them off. Neil Jacobs, acting programme director for digital infrastructure at Jisc, the UK higher education IT consortium, pointed to considerable activity around some repositories. The University of Glasgow facility, for example, records 20,000 downloads every month.
He said the growth of repositories' contents could be boosted by projects to automate the deposit of papers into repositories after the specified embargo period - typically six months - has passed.
Other defenders point to repositories' suitability for hosting non-traditional academic outputs such as videos and datasets, and for helping universities to showcase their research and prepare for research assessment programmes.
No quality control, publishers warn
Michael Mabe, chief executive of the International Association of Scientific, Medical and Technical Publishers and a visiting professor in information science at University College London, said that publishers were increasingly uncomfortable with the threat posed by repositories.
He noted that even cutting-edge science papers see only half of their total downloads within six months, meaning that publishers who allow papers to enter a repository after that period are giving away large numbers of downloads.
Professor Mabe also feared that once repositories contained a significant volume of material, libraries might no longer feel the need to subscribe to journals at all, leading to their demise.
This would be disastrous, he argued, because repositories are essentially "electronic buckets" with no quality control. He also expressed doubts that the academy would be able to successfully introduce peer review to such repositories, partly because it would be difficult to attract reviewers who had no "brand allegiance" to the repositories.
Mr Pressler agreed that repositories in their current form are no substitute for high-impact journals, as association with such journals - as an author or editor - conferred considerable professional benefits.
He said the open-access movement had been misguided to push for institution-based repositories when academic communities were built around disciplines. Instead, he argued, advocates should have tried much earlier to convert research funders to the open-access cause and ask them to organise funding and peer review for repositories in their subject areas.
But Mr Pressler argued that it was not too late to recast the existing infrastructure in the form of international subject repositories or e-journals that would replicate everything that existing journals did.
Cameron Neylon, an academic editor at peer-reviewed open-access journal PLoS ONE, agreed that getting the most out of repositories would involve thinking of them as a mechanism for publishing as well as archiving. But he said that enthusiasts were held back by their "terror" of a publishers' backlash.
Professor Mabe pointed to the history of public sector IT projects running over time and over budget and accused librarians of employing "voodoo economics" to demonstrate repositories' cost effectiveness.
But Mr Pressler said he was very surprised that it had not occurred to cash-strapped governments and funders that huge savings could be made by taking publishers' profit margins out of university budgets.
He argued that however loudly academics and publishers might protest, it could be easily achieved with enough political will. "If the combined journal invoices of the Russell Group were redirected to create a properly governed, peer-reviewed open-access landscape, it could be achieved collaboratively," he said.