Exactly!Mortality rates vary from trial to trial because the...

Exactly!
Mortality rates vary from trial to trial because the patients in each trial tend to be different, got their injuries in different ways etc. This is why you always have to have a placebo, because this controls for the different baseline characteristics of the patients in the trial. If the overall mortality rates are substantially lower in this trial, then as I said before, it could be due to just the active arm i.e. NNZ2256 having much lower mortality, with the placebo arm having 'average' mortality, or it could be that in this study everyone had lower mortality due to the mode of injury or the overall standard of care being superior now compared to a few years ago. In the latter situation it makes it more difficult to prove the efficacy of the active drug: if NNZ2256 reduced mortality by 10% (which would be very good given that aspirin only reduces mortality by 4% in acute MI) then if 50 patients die in the placebo arm, 45 would die in the NNZ arm and depending on overall numbers that difference could be statistically significant. However, if only 10 patients die in the placebo arm, 9 would die in the NNZ arm and it's highly unlikely that would ever be statistically significant.
So, yes it's good that there are no increased SAEs in this study, but overall reduced mortality may actually work against NNZ2256 if the standard of care arm has low mortality. This, in fact, is what happened to glypromate. It's generally best to conduct trials in populations which a high incidence of adverse outcomes.