Well, yeah, there are other irrelevant things about reviewing that I also have the ability to bring up but my point is that to say that the "benefit" of reviewing is for the purpose of networking is misinformed.
The authors most likely won’t (can’t) ever know that I agreed to review their paper; it’s more of a good citizen affair. Sometimes, afterwards, it will become apparent that a particular referee was someone familiar. I would like to think that the original poster was imagining something along those lines, but your take is probably closer to the truth. Oh well.