I stumbled across an old article from back when Google ruffled many feathers with the roll-out of the "shared endorsement" thing. I think the suggestion of using the "glass" marketing method is extremely interesting, and I'd love to see somebody - FB, Apple, Microsoft, Google, whoever - try it out.
It would be fascinating to see if the "social market" could be leveraged in such a way, and if the "marketing market" could manage to provide sufficient positive experience and incentives to give the voluntary beta testers cause to signal influencers to onboard. I even suspect that in general people would be ethically okay with defaulting "on" for NEW accounts, with disclosure, it's the opting-in of extant accounts without consent that seems to push the creepy exploitation button, not to mention how people get weary of having to go update their privacy settings again and again. If there's REALLY an ACTUAL value-add or reason for such things, the social is very capable of getting people on board.
Even the much--hated Facebook has a value - the mass of people that people actually know are there, people were able to connect with people from other parts of their lives that they'd lost touch with. The maligned game invites were also opt-in, people voluntarily spammed their friends until the social decided that was Rude (and people got bored of most of the games) and that's largely Over With for anyone with a social clue now (props to FB for offering tools to help handle that for those willing to make the effort, too) - but people are still on FB, and like actually so, not like inflated by Chromecast screen savers not even active on the screen there.
So it seems like it COULD work as a social-market opt-in... and one might even go so far as to posit that if a "feature" is the sort of thing a company is 100% sure people wouldn't opt in to even with carrots/transparency/lack of lock-in, it might not be that great of an innovation after all. #READMORE
(I'm also skeptical that scrolling-past should count as a "view" on a text-post, given that "view" is only a couple lines. I could understand a photo view counting in scroll-past, since photos are giant and almost fully displayed, but oddly enough it seems to be claimed that DOESN'T count. And maybe "served" is a more accurate term than "viewed" given the apparent metrics involved.)http://adage.com/article/digitalnext/google-played-terms-service-announcement-wrong/244887/