I believe it's a mistake to characterize people who don't "get" online worlds as somehow defective. They're not. Not appreciating online play does not mean that there is something wrong with these people intellectually, ethically, emotionally, politically, or otherwise -- all it means is that they have a different understanding of what's important than people whose primary goal is something other than security.
Nor is it correct to say that this difference in world-goals is a matter of learned habits, or a reaction to some "trauma" experienced early in life, or any other such environmental phenomenon. People are just born different. They pop out of the womb with different fundamental motivations, and to a great extent they retain those different motivations throughout their lives.
A study of the behavior toward strangers of newborn babies (Thomas & Chess, 1977) characterized newborns as outgoing, reserved, or "slow-to-warm-up." Some babies would coo and express curiosity at being held by new people; others would cry if they noticed a stranger; others would express concern but would accept the stranger as non-threatening after a time in which nothing bad happened.
What's interesting is that these attitudes toward security are apparently retained throughout one's life, and underlie decision-making regardless of most life experiences. They appear to be based on innate preferences... which means that we ought to view those preferences non-judgementally.
(That's not to say that actual behavior should be immune to judgement, just that the basic motivations native to individuals are not "wrong" merely for being different.)
As for the specific motivations themselves, there are any number of reasonable categorizations, ranging from Myers-Briggs and Keirseian temperament theory to "Big Five" models to the description (by one of the DSM-IV's creators) of pathological behavior as a hyperexcess of one or two otherwise acceptable motivations. The main thing to recognize about all these models is that they begin with the modern assumption that "different" doesn't equal "wrong" where motivations are concerned.
So any of these models can work for understanding the behavior of people generally and gamers specifically. None of them are perfect ("it's only a model"), but perfection isn't required -- they just need to explain human behavior to a useful degree. I've found the models I've named generally good enough (because they're models based on empirical data, and not data generated to support a theory), but others may work as well.
Speaking for myself, I happen to prefer the model David Keirsey developed, with a few minor tweaks based on my own experience and analysis. To express it in terms appropriate for this discussion, most people are guided primarily by one of four motivations:
(The names given for each of the four temperaments are Keirsey's; the descriptions are mine.)
Myers-Briggs statistics indicate that there are a lot more Guardians and Artisans than Idealists and Rationals -- that is, the general population contains a lot more folks who are concerned with the physical, concrete, external world than with conceptual, abstract, internal phenomena. Most people are content -- in fact, prefer -- to concern themselves with what is "real": building things, making money, shaking hands, acquiring possessions, manipulating objects and people. To these folks, even the word "game" has certain undesirable connotations of something adults just don't do, or something not meant to be taken seriously.
OK, so some people aren't interested in -- and live perfectly satisfying lives without -- playing computer games. So? How is that a problem for those of us who do find satisfaction in playing (and thinking about the playing of) games? Why should we consider others broken because they aren't like us?
Of course it's also true that Guardians and Artisans are wrong to consider gamers somehow defective because we don't share their externals-directed motivation. If someone doesn't want to play games themselves, that's fine, but if they go beyond this to attacking the value of gameplay (in moderation) to others, or even trying to prevent others from engaging in appropriate play, that's not fine and it deserves to be challenged.
But it's not a mistake we should copy. If someone is wrong to deny the value of play, let's not emulate them by questioning the value of security. If we don't let others be different, why should we expect them to let us be different?
If we don't respect their motivations, why should they respect ours?