There is a very subtle distinction here, no? What you're talking about is relevancy but the criticism levied against Facebook, CA, et al, is that it's not necessarily relevant per se, but manipulating fears and emotions of people based on easily identified characteristics.

If this was purely relevant then you'd expect Facebook and Google to serve you content that challenged your opinion, rather than vying for your approval. As it stands, the technology understands you like certain things and does whatever it can to hide things it assumes you won't like.