Your idea about latency in the context of belief propagation seems to have potential (and looks novel, as far as I can tell).
Up voted article because of this.
Up voted article because of this.