By nature (ahem), editors of these journals are responsible for what goes in their pages. Decisions of what to run have some motive - you can call them "political," or "profit-driven." But this has always been there. If an editor thinks your research isn't in the cutting edge, it won't get published. There used to be an override mechanism (in PNAS, Academy members could recommend others' papers for review).

If you need anecdotal examples, I'm sure we can provide some :)