Tolkien disliked allegory so intensely because he felt it was too didactic. It leaves no possibility that any other levels of meaning in the work could exist. Tolkien understood the artist, created in God’s image, to be a “sub-creator” — producing a work of the imagination that functioned best when it followed God’s own complex action of creation.
To do this most successfully, a complete alternative world had to be created in which the work of redemption could be played out within its own consistent and logical constraints. It was not enough to create a world with symbolic pointers to Jesus Christ and the cross; that world would have to have a whole history and unique inner dynamic that would incarnate the universal truths in a totally fresh way.
The difference between Narnia and Middle Earth points to the underlying difference between the imagination of Lewis the Protestant and Tolkien the Catholic. For the Protestant, truth is essentially dialectical. It consists of abstract propositions to be stated, argued, and affirmed or denied.
For the Catholic, Truth, while it may be argued dialectically, is essentially something not to be argued but experienced. The Truth is always linked with the mystery of the incarnation, and is therefore something to be encountered.
Many Protestants will argue, for instance, that God’s primary revelation is Sacred Scripture, while Catholics maintain that God’s primary revelation is Jesus Christ. That Lewis produced works that were profound, worthy, and beautiful, but less than fully incarnational, while Tolkien produced a masterpiece that incarnated the same truths in a complete, subtle, and mysterious way reflects the deeper theological differences that remained between the two men.
Longenecker’s argument that didactic nature of Lewis’ fiction stems from his Protestant faith is dubious for several reasons. For one, the argument isolates the Chronicles of Narnia as paradigmatic of all of Lewis’ fiction. While Lewis is often didactic (especially in Perelandra), no claims about his literary imagination are complete without reference to ‘Til We Have Faces, Lewis’ best piece of fiction. ‘Til We Have Faces is a rich, intricate examination of the nature of love, and no more didactic (I would argue) than Lord of the Rings. In its depth, gravity, and style, ‘Til We Have Faces is, in fact, a worthy equal to Tolkien’s work.
Secondly, the fiction of G.K. Chesterton–no slouch of a Roman Catholic himself–is even more didactic than Lewis’s. At points, his stories serve only as backdrops for his characters’ always amusing and edifying speeches. For Longenecker’s argument to stand, he would have to agree that Chesterton’s imagination was more Protestant than Catholic–a thought, I’m sure, which Chesterton would himself reject, and which would be difficult for any reader of Chesterton to sustain seriously for long.
To conclude, then, while I would grant the superiority of Tolkien’s Lord of the Rings over the Chronicles of Narnia, the idea that this is due to their respective theological systems is tendentious at best. Longenecker passes over more mundane, boring, and probably accurate explanations–Lewis wrote many things besides the Chronicles of Narnia, while Middle Earth was Tolkien’s central focus–and reaches beyond where the evidence (at least as he presents it) leads.