I like the idea of green open access as a compromise among different perspectives.
When I’m working with collaborators and discussing where and how to publish, the impact factor of the journal matters a lot. (I may not like that fact…but I do have to deal with it.) Having the option to self-archive, and knowing enough about the copyright terms of important journals to be able to preserve that option to self-archive, is really valuable. Researchers are increasingly inclined to stick their necks out and decide to only publish open access, which is awesome, but many have to choose between their commitment to open access and their desire to build a career and a CV that will support that. I like being able to say “Okay, let’s publish in that prestigious [closed] journal, but let’s also release the data set, and let’s deposit the paper in [appropriate repository] according to [prestigious journal]'s terms and conditions.”
Embargo periods, while frustrating, offer another way to persuade a collaborator who might be nervous about releasing their work openly. My thesis advisor was amenable to letting me release my thesis openly as long as I kept it under embargo for a year, which was a good compromise.
edited to add: Green open access may also provide better data longevity – an institutional or field-specific repository may be less subject to winds of change that might cause gold access sources to open and close over time. Openness without sustainability doesn’t help anyone in the long term.