What *should* the publishers lobby for?

The Research Works Act is a very poor move by traditional publishers.

Publishers will come out strongest if they side with the future: the future is immediately open, reusable, all-value-added versions of research results and peer reviewed publications.

Sure, they should lobby to shape government policy to their favour.  That’s their right and maybe even their responsibility to their stakeholders.

What *should* traditional publishers fight for, to stay on the right side of both history and their balance sheets?

1. Time.  They should insist that any federal mandate that requires the article-of-record be made openly, immediately available does not take effect for a year or two, to give themselves time to change their business models (to author/funder pay-on-publish or pay-on-submit or some other method, thereby saving their companies and jobs).
2. Access to publication funds for federally-funded authors.  Publication costs are already available to NSF and NIH awardees as budget line items in grants.  I don’t know if all other federally-funded investigators have access to author-pays grant money… if not, publishers should argue that access to these resources must be a condition of a mandate.  There must be a creative way to redirect money which payed for university library subscriptions into university OA funds or federally-disseminated research distribution reimbursement (has anyone proposed such an approach yet?)…. publishers should lobby for this.
3. Measurement of the impact of the papers they publish.  When research papers are openly distributed, redistributed, deconstructed, and mashed up it becomes much harder for publishers to understand (and therefore brag about, and then capitalize on with higher publication charges) the impact their publications have had vis-a-vis their competitors.  Publishers could insist that all federal hosting services report back usage stats (as PubMed Central does), and lobby for requiring a manner of attribution that facilitates easy and robust impact tracking (beyond just mention or citation).

I’m not saying I think they should get all of these things, necessarily… but this would be a much more constructive, productive and supportable stance.

What do you think they should fight for?