All studies are skewed. It's all interpretation of the available data. Again, there's no way to make anything 100% safe. There is no way to prevent injuries in every situation. The helmet is a reasonable step in the direction of safety. A c-collar wouldn't be practical for us to use, but perhaps you're onto something. How about those big ol' neck pad things racers and stunt drivers wear? While we're at it, kevlar vests have saved the lives of many vehicle crewmen in accidents because they took a hit from a sharp edge during a tank rollover or a helicopter crash. Of course you've got seatbelts too, but we've beaten that horse to death. If anything the helmet is still less than adequate protection when in the patient compartment, but it's a step towards safety.
As for the seated or supine patient, they're supposed to be strapped in according to our local protocols. That's feasible with the included straps and belts in most situations.
Remodeling compartments has been occurring. Look at the older rigs in service (y'know, the steel box with the wood paneling) and this year's models. More rails, more padding, less protruding objects.
I do agree that more data has should be taken into account to study common injuries that occur in the back of the ambulance. The one study approach primarily benefits the politician much more than it benefits the end user.