His final chapter in that book is devoted to the late period of Freud when he turned his attention to the Future of an Illusion, a book, as I've often noted, Freud himself denounced almost as soon as it was published as "my worst book!" Lear's assessment of Freud in this book and other works is very helpful. He begins by noting--as others have--that Freud very much wanted to situate himself as an Enlightenment rationalist par excellence, and as a successor to Darwin. Such desires led him to some serious mistakes in writing about "religion." As Lear puts it with great clarity:
in the name of analyzing the fantasy underlying religious belief, Freud participated in his own fantasy of inevitable historical progress, which included secularization as a hallmark of that progress. There is reason to think that this closed down Freud's curiosity: he was disposed to see religious commitment as historically retrogressive. If he could find a kernel of wishfulness in that commitment that was sufficient; it was as though there was nothing more to look for. As a result, Freud blinded himself to the possible complexity of religious belief (204).This very much accords with my own read of Future. It is insufficiently curious, ideologically pre-determined, and in some ways also very lazy: he never bothers to move beyond sweeping and sophomoric generalizations to investigate the depth and details of what he denounces too facilely. He also represents Christianity in particular as nothing more than a cult of the "primal murder" of the father by the son, a notion that is laughable on its face as a few seconds reading the New Testament will reveal.
As Lear puts it, Freud fails to do to his own analysis what he readily applies elsewhere. But he also makes historiographical mistakes--though, admittedly, these were not uncommon at the turn of the century, when for some time many thinkers had been predicting massive and unrelenting secularization so that, by century's end, "religion" would largely have disappeared. We now know what a crock those predictions were and are. (MacIntyre's 1967 lectures on this are still useful.)