These studies have attachment as the independent variable and well-being later in life as the dependent variable. While children with secure attachment tend to be better off later as the results of the experiments state, correlation does not equal causataion. Thus well-being being a direct factor of attachment style would be a possible implication, but saying that it is a direct cause and effect relationship would be false. For example another factor such as exposure to different types of people early in life could be another factor. Even if the child is securely attached to his/her parent they may be exposed to othe people who are not as interested in the child's best interests, and thus a child with a secure attachment to his/her parent may grow to be cautious and wary.
This is a nice point on why the conclusion from the research is not valid based on statistical method, but the alternate explanation provided isn't really an alternate explanation as to why there could be a positive correlation but and explanation of how the correlation could be a negative one. So I don't think it really answers the question but offers a few good points.
