If we possessed only a handful of manuscripts for the Old and New Testaments, it would indeed be difficult to reconstruct the reading of the original. However, the large amount of textual evidence for the Old and New Testaments, while increasing the number of the textual variants, makes it easier for us to reconstruct the reading of the original text. Rather than undermining our confidence in the Bible, these variants make it possible for us to determine, with near-perfect accuracy, what God originally communicated in His Word. (p. 110)
Nick's response included the following measured words of wisdom:
At best we can know with precision what the earliest extant text said; but we can only guess with varying degrees of probability what the original text said. So when the Alands say that the original reading is present somewhere in all the variant readings, and this claim is repeated by men like Dan Wallace or James White, it’s at best wishful thinking. It’s a statement that will always need to be qualified with “I think” rather than “I know.” This isn’t to say that we can’t guess with a very high degree of probability, but if we’re honest, in the end we’ll always have to succumb to some kind of textual agnosticism with respect to the originals.
His comments are well taken, although it's important to point out that the reading which is ultimately determined to stand closest to the original is not always that of the earliest extant manuscript (and I'm not exactly sure what "textual agnosticism" means, although it sounds cool!). But I've always shared his incredulity over the Alands' absolute certainty that the original reading exists within every set of variants, even if this is true in a large majority of cases. The ending of the Gospel of Mark represents one notable instance where some scholars believe that none of the extant variants are original; Rudolf Bultmann, for example, argued that the gospel must have initially concluded with a series of resurrection appearances in Galilee. And of course (and I've repeated this so often that I feel a bit like a broken record) we have very, very few New Testament manuscript witnesses from the most dynamic period in the life of any document: the first century after its composition. So while we can feel relatively confident about the reconstructed text of the New Testament, we should bear in mind the limitations of that reconstruction.
As for the argument that variants make the determination of God's Word even more certain, this is largely lost on me. I'm not sure how particularly difficult critical decisions (e.g., Romans 5:1, which I've discussed here) would support such a view.