hardburn,
Well, technically it shows a set of conditions in which failed local realism doesn't lead to faster than speed of light communication. Taking it further, it shows that it would statistically be impossible to differentiate changes to state as intentional communication or side-effects of observation itself. This seems to imply that you might be able to communicate this way but it would definately be unreliable.

To be honest, I don't think we have scratched the surface on understanding this stuff and hence - not holding my breath.