This is not open access. The Berlin declaration states it quite clearly. It is about usage rights.[0]

Publishers really love co-opting the OA movement. Scheming up increasingly convoluted colour OA categories. Or deluding the concept with campaigns like this.

[0] https://openaccess.mpg.de/Berlin-Declaration

So they are doing 'Open Access Week', I tried to figure out why, and why this week, which led to me this page https://royalsociety.org/journals/free-content/ and then to here:

http://www.openaccessweek.org/page/about

The contents might already be on SciHub and other sites, but I can already see the bandwidth spike they're going to receive from doing this... from those sites wanting to complete their collection, among others.

Why not be open access always if it can be done for a week?

The Royal Society is a non-profit organisation - the income is used mostly to fund research, but also to fund scientific collaboration, and on public policy to try and ensure governments are making decisions that are not completely ignorant of the scientific evidence (of course they can then ignore that evidence).

From the RS financial statements, the income from its publishing side raises about £7.5m per year, and the publishing expenditure costs are about £3.5m per year. So that's around £4m/year that is invested in science that otherwise wouldn't be.

I think many RS fellows would like to see the RS journals become open access all the time. But it would mean that less research could be funded.

Disclaimer: I'm a fellow, but I don't have any insight into future plans on this front.

Put all the back catalogue of research into a zip file and throw it out there via bit torrent.

Let the downloaders become the sharing infrastructure.

Much of the back catalogue is already available open access direct from the Royal Society:

Most of our oldest content is now freely available, specifically, all papers older than 70 years. In addition, papers published between 10 years ago and either 12 months ago (biological sciences) or 24 months ago (physical sciences) are freely available. For Biographical Memoirs all issues are now freely available, apart from the most recent issue.

That's odd. What about the papers published between 70 and 10 years ago?

Ancient organisations like the Royal Society have arcane rules and oddities. This is likely one of them.

Sci-hub is already available, no need to send zip files around for distribution

Sounds like a different funding model is needed. Retarding or advancing the progress of humanity is the bigger picture here.

The £4m was mostly spent by other scientists, so there's no net benefit to science research spending.

Is £4M of research funding worth more than the impact removing the friction and cost of accessing past articles would have on research (in terms of both cost and speed)?

Because if you're not the customer you're the product. For-profit publishers have abused the reader-pays model into disrepute, but author-pays and sponsor-pays is an awful way to run a journal.

Money