I think, having lived as a member of that community as well, that the comparisons with Type I diabetes ARE valid.
Most of those diabetics that prefer a more laissez-faire approach to management are "fine" for a long time. They, too, THINK that they are choosing to "not let this thing get to me" or to "live my life anyway."
I think that the underlying psychology is quite similar. It's a form of denial, in any case. That part is comparable. But only to a point, because at least everyone diagnosed with Type I diabetes is actually impaired in terms of insulin production.
The compounding factor here, IMO, is that some of that group really doesn't have the same disorder. That is, thanks to bad diagnostic criteria, there are probably a fair number of people who are labeled as "FA" and even prescribed injectors-- who really do not have food allergies at all. Then there are the majority of food allergic people who have relatively high thresholds in this mix, as well.
My DH has a very high threshold for walnuts. <shrug> Is he wrong to not carry epinephrine? Probably not, realistically. The odds of him actually ingesting enough to cause him to anaphylax are pretty much zero. He's one of those people who can smell/taste the allergen and he knows the likely hidden sources well enough that he just avoids them completely unless he is 100% confident in the ingredients. He's never actually anaphylaxed, in spite of his many allergies to a variety of things. He may not have the genetic potential for it, in all honesty. There's no diagnostic way, at least currently, to identify those people who are NEVER (well, probably never) going to anaphylax because their immune systems aren't made to do that. (This is different than component testing a la peanut seed-storage proteins, btw.)
My DD and I are another story entirely; both of us have ample and robust anaphylaxis history. We both carry epinephrine.
All three of us live in such a way that we're not taking excessive risk-- but DH appears almost unlimited, while I'm definitely more limited in my 'dining out' options, though not much otherwise... and DD is one of the people who looks "crazy" in terms of avoidance.
Until the public is capable of understanding the nuanced picture that emerges just in looking at the three FA individuals in my own family, there will be problems in the way that the <1 to 5% of people who are/must be as careful as my DD are percieved.
Heck, physicians aren't always sure what to make of the three of us and the story that our anecdotal experience has to tell.
Point being-- how many people in that HMO sample were diagnosed with a PA on the basis of a high RAST alone? Does having an elevated RAST to peanut mean that they all NEED autoinjectors? Probably not. Estimates are that at least some percentage (20? 30?) of them aren't allergic at all to begin with, and that some other fraction have an allergy to one of the components that doesn't put them at risk of a systemic reaction... and then there are the people like my DH, who would have to literally EAT pb in order to really react.
It's a very complicated picture. The only really clear thing is that kids who have already had very severe reactions or react to very small traces are likely to be at highest risk. Beyond that, it's mostly speculation and the picture that emerges, mirage-like, shifts a little with every research study.