It is interesting to note that the OP posted back in 2011 and the debate goes on still.
The debate started WAY before 2011, and it shows no signs of slowing down. After quite a few years of staying on top of the research related to EMS airway management and debating it ad nasueam, I have come to a few conclusions as to why the debate continues.
Suffice it to say that there is no other intervention that offers as much risk to the patient with as little demonstrated benefit - yet somehow continues to be considered "standard of care" - as prehospital endotracheal intubation.
The article cited by the OP came from PUBMED, a highly reliable source for EBP. However, in the article cited, it did not specify which locale, if any, or if it was a nationwide study. With the low sample population, I would think it would be relegated to a specific urban/suburban area vice nationwide.
PubMed indexes practically every article related to medicine that is published in journals written in English. So the quality and strength of the evidence presented in each article varies dramatically by study design and execution.
This study was done in Orlando over an eight month period in 1997. It is one of the rare prospective studies done in the US on prehospital intubation. With a sample size of 108, it's pretty legit in size, though you are correct in pointing out that it is limited to a single geographic location.
The reason for the high rates of unrecognized esophageal intubation in the study appears to be largely related to lack of use of Etc02 monitoring.
I have engaged in debates on different forums about ET intubation, and some of the replies went to the lack of people surviving cardiac arrest neurologically intact who were intubated. I question the validity of this argument in that an ETT is not the sole arbiter of patient survival. Length of down time (anoxia), length of delay in ROSC, comorbidities, etc., all lead into whether or not the patient survives, neurologically intact or not.
Well of course it isn't the sole arbiter but age, downtime, and co-morbidities are pretty easily controlled for in these studies. The investigators generally group the patients by those traits and run stats separately on each group. So that isn't really a problem. The evidence on how ETI affects survival from cardiac arrest is inconclusive. Some analyses show a benefit, some show that it doesn't help or even negatively correlates with survival. Personally I think we put too much emphasis on trying to bring back the dead so I tend not to pay a heck of a lot of attention to studies looking at different airway management techniques in cardiac arrest.
Where the rubber really meets the road is in looking at field RSI. There are a handful of retrospective studies on field RSI over the past 20 years and none of them have shown a clear benefit, while many have actually shown harm.
There has only been one large, well-done prospective study and that was the Bernard study done in Australia and was also very unconvincing.
Misplaced tubes should be a rarity, not commonplace.
Since Etc02 has become common in the field, misplaced tubes really don't happen that often. Most studies on field RSI show pretty high rates of success with tube placement. The problem as it relates to benefit appears not to be getting the tube, but all the other details.