Trauma, unsurprisingly, derives no benefit from ALS in cardiac arrest. They even found a pooled OR of 0.89 for survival with ALS, but the CI just barely crosses 1.
But, contrary to the two most recently published prospective trials, their meta-analysis of non-trauma arrest still shows a survival benefit for ALS. They do include a few trials from before AEDs were available in BLS in 1995, but it still doesn't explain the entire benefit. They also cite a few studies in which a physician is part of the paramedic team, which may mean there's more to ALS than AHA ACLS, so that might be a bit of a confounder. Hard to know what to make of this data, considering the lack of demonstrable benefit from ACLS medications and the decreased survival of patients intubated in the field in out-of-hospital arrest.
My take is still that cardiac arrest, for the moment, is still a place where significant out-of-hospital resource investment is low yield, and CPR and AED is all they need en route to the ED.