Other than BMJ's high profile, what makes BMJ's change interesting is that there's a print journal–and it's becoming a selection of BMJ's published articles rather than the totality. Assuming nobody tries to make a branding distinction between “accepted but not part of the print record” and “part of the print record” this strikes me as a useful innovation.

[I'm in the process of building a cluster on open access at the PALINET Leadership Network, so I've been thinking about this more than usual.]