Here's the lowdown on generalizations.
Generalizations are fine if we are interested in a sociological sense in how certain groups tend to behave in certain ways. However, if they are being used to write off a group's opinion, then they are fallacious.
As an example of how generalizations can easily be used to write off others, I can cite the decades of research indicating that: religiosity is inversely correlated with IQ, test scores, GPA, and educational attainment more generally, plus the data showing that more intelligent children end up being less intelligent later in life, and that extremely few eminent scientists -- especially the physicists who study the celestial mechanisms Christians constantly tell us must have a religious explanation -- are religious. This holds within the U.S. and across a wide swathe of countries. The correlations are strong enough such that I can pick out a random Christian in this country and say with a high degree of confidence that they are less intelligent than a random irreligious person I pick out.
This is an interesting generalization to discuss and indeed, I have a theory as to why less intelligent people are attracted to the ideas of Christianity. But under no circumstances do I use this generalization to write off Christian opinion. One must still argue the specific propositions at hand and listen to the other side, genuinely trying to understand them before evaluating their arguments. I still must have good arguments against the supposed historical evidence for Jesus' miracles, for the supposed necessity of God's existence, and so forth. Generalizations about Christians don't relieve me of the burden of actually having a defensible worldview.
So too with Christians. If atheists have a tendency to be insecure, less happy, liberal douchebags (I haven't looked at data corroborating any of this so I don't know), then this will be an interesting sociological fact but it will not do anything to help Christians make their ideas more defensible.