[rab] Count yourself lucky that the editors are sending you such good papers. In my field at least, rejection rates in the top journals run 90-95% or so. I certainly have to recommend rejection much more often than acceptance.
[rab] It's positively a public service to see that rubbish doesn't reach print. In the past I've rejected several papers for triviality, several for technical slipshoddery, one for outright crackpottery, and one for unwittingly reproducing (but not as well) some of the results from a certain fundamental paper that hardly anyone has read but everyone cites (including the authors of the rejected paper). I assume your refereeing process is anonymous?