If you're an ER doc, what makes you qualified to criticize their statistics?
1) It is impressive that they managed to recruit so many patients for such a complicated study, and I don't think their sample size is that small in the first place.
2) If you do a power analysis, it won't tell you whether the 18% difference is "real" after you've already established statistical significance.
3) Regarding intravenous thrombolytics: So the 28% vs 23% is "not all that different". But could it make a difference to a few patients? The "lousy p-value" (p=.043) indicates a significant difference at alpha=.05, so I have no idea why you call it lousy.
Your post basically showed that ER docs will use the year of stats classes that they took in school to interpret research any damn way they want to.