I agree that this is a big question. What are the divisions between manipulation, persuasion, coercion, influence, information, and education? Intent? Whether or not we agree with the position? I know that this can come off as flip, but it's not at all meant to.
In the case of the poster, what if there were two posters, an additional one for the local university. Average recruitment goes up 3%, and applications for the local university go up 3%. Does that change our perspective on what happened and the students' choices?
In this new scenario, students read the signs, the messages are brought to their attention, and of all the possible things they might do, the odds of choosing one of these two things increased some degree. In other words, the external forces in their environment have shaped their future state (choice).
I'd be curious to see a table of interventions, and the degree they pull out the entropy of what a person might do next. We might then make some categories, like you mentioned above.
When we appreciate the power of manipulation, for better or worse, perhaps we'll give it more serious consideration. This ought to undergird our thinking and decision making, whatever we may CHOOSE. ;)
In the lunchroom example, so long as lots of other folks are putting posters on the wall, and presumably some kids don't look at the lunchroom posters, I don't have a problem with either the college or the military changing behavior that way. There were many alternatives, and the poster wasn't forced on anyone. The influence we're seeing now is expensive (much higher proportional impact if you can pay), and effectively guaranteed (everyone is tuned to the same channel or two).
I hope we don't have to analyze individual interventions, I'd much rather just break up the channels, or increase the diversity of the programming. Mostly because I'm not sure how to (and I'm not sure we should) answer your question.
Yes! In all organizations, money and power buy influence. The internet briefly changed this, but it has reasserted itself.
What do we expect the powerful to do with their influence then? Rules for rulers says they must use it to increase their power for any other decision reduces their evolutionary fitness and long term survival.
Therefore in the long run altruism is strongly selected against and we would expect to see less of it.
For example, Suppose on average 1% signup for the military in high school. Next suppose we add an intervention, we add a sign in the lunchroom for military recruiting (touching on all the right emotions). The average in these schools goes up to 3%! We say, these additional recruits CHOSE to join, but we know they wouldn't have had we not added the sign.
It's in this way that I say that realtime voting doesn't solve the problem, because it doesn't look holistically enough, namely at influence. It's not just at the ballot box, but in your daily life. The question then becomes, who can influence and how much?