Is the backlog of a journal correlated to the difficulty of submission (with acceptance) to a journal?
I would imagine that a journal with high backlog could be more selective in publishing journals (if they have so many journals to choose from).
At least in mathematics, this seems anecdotally true. For instance, Annals of Mathematics has one of the largest backlogs (and is also notoriously among the more selective of math journals).
On the other hand, I suppose it could be possible that backlog is merely a measure of how slow or fast the journal is, and could have nothing to do with difficulty of getting accepted submission.