My problem isn't that academics do that, it's that some, maybe the large majority thereof, do it with nothing to show for it. Seriously, I don't remember reading any relevant paper from a Thomson-endorsed journal or conference in the last decade, and trust me, I've read quite a lot of them in the first half of it. Maybe it's just my bias towards the computing sciences and they're doing quite fine in natural sciences or the various fields of engineering, I don't know.

The (perhaps honest) counter to this would be that likely institutions such as ACM and IEEE have already been corrupted by idiots -- frankly, I myself think that's just how these things usually go. On the other hand, I've met a few IEEE fellows in my field who for the better or worse manage to move things, albeit slowly. The alternative to so-called standards is everybody rolling their own and no one getting to talk to anyone else, which is (IMO) way worse.