Yes, rendering a selection makes more sense. I was objecting against a "render channel" function because you want to render one channel, the next person two, then another person five channels... I thought there was a request for this once, but I can't find it anymore, so you may create a new one on the issue tracker.