JISC has released "Scoping Study for a Registry of Electronic Journals That Indicates Where They Are Archived."
Here's an excerpt from the "Executive Summary":
The research and especially the interviews have confirmed the assumption behind the project that there is a need for more information, and more easily accessible information, about where e-journals are archived. However, what has also emerged strongly is that this issue cannot be considered in isolation, either from the overall context of relationships within the scholarly communication system, nor from other initiatives being undertaken to improve information flows e.g. in relation to the transfer of journal titles between publishers. . . .
Librarians felt that they were most likely to consult a registry in situations where they were considering taking out or renewing a subscription; considering cancellation of a print subscription in favour of an e-only subscription; contemplating relocating or discarding print holdings. The vast majority of potential users of such a registry would be library staff in university and national libraries, though organisations licensing e-journals on behalf of the library community would also be likely to use the registry to check compliance with licence conditions.
One of the key benefits of a registry is perceived to be the exposure of gaps in archive provision. This was identified by all types of stakeholder: librarians would want to be alerted to risks to any of their holdings; publishers who are making provision would like to see their efforts recognised and pressure placed on publishers who are not making satisfactory arrangements; archive organisations would also benefit as that effect fed through to more demand for their services.
The drawbacks to a registry as a solution to the acknowledged information gap were mainly seen as ones of practicality (keeping the information accurate and up to date), trust (especially whether a national solution is appropriate, and conversely whether an international solution is feasible) and sustainability of the funding model. Other solutions were suggested, mainly involving either WorldCat or ERM vendors such as Serials Solutions. The latter were also suggested as a complementary part of a solution involving, but not limited to, a registry.