I'm not aware of any papers showing a reduction in mortality.
Early on in the study of MI, back before we even gave aspirin, lots of patients would have VT while in the hospital. Beta blockers would reduce these arrhythmias.
I often thought of the concept behind beta blockers as decreasing workload. Apparently, this wasn't the case. They were being used as an anti-arrhythmic.
There have been some suggestions recently that the use of beta blockers in patients who later develop cardiogenic shock is linked to increased mortality but I do not believe causality was conclusively shown. The big problem, of course, is we don't necessarily know who will develop cardiogenic shock in the next several days when we give BBs up front.
So... potential risk, no real gain. Probably shouldn't be doing it blindly for everyone with chest pain. This seems to have changed only recently in my experience. Even in my last year of residency (I'm starting my second year out now) we would give lopressor to everyone. I gave it for the first time in over a year (in this context) last shift but only because of specific indications (hyperdynamic, hypertensive with dynamic ST changes).
Sorry I can't serve up the specific references, I'm about to head to work. You'll have to consider this anecdotal information.
Take care,
Jeff