I tried to access the articles, but they are not yet posted to sciencedirect. It's tough to tell from the abstracts how exactly they calculated survival.
The first study is out of the UK and the 2nd out of Australia. Both countries have a low burden of ballistic trauma and as mentioned previously, a single stab wound is typically associated with the best survival (relatively speaking), and both countries, as far as I know, see more stabbings at least insofar as proportion of penetrating trauma. Also, at least in the UK, it is commonplace to have physicians not only respond, but to perform resuscitative thoracotomy for select cases of traumatic arrest. London HEMS has pretty much shown that their only survivors are those with a single cardiac stab wound. I know that in many area of Australia, it is also common to have HEMS staffed with a physician, but I do not know if thoracotomy is a procedure that they perform routinely for traumatic arrest.
I'm a resident at a trauma center that has ~2000 activations for which nearly 40-50% are penetrating. The trauma service performs approx 50 ED thoracotomies per year and are very liberal (penetrating wound below the neck and no pulse, you get opened). A very small number survive. I have seen short term survivors, but I personally have not cared for a patient that survived to D/C that I know of, but the trauma surgeons affirm that there are survivors.
Anyhow, I would not be surprised if the numbers are much lower when you look at all patients who at one point in the prehospital setting were pulseless 2/2 trauma. I suspect that 5% and 7.5% represents a subgroup of the all-comers.