Summary: An examination of the three earliest extant copies of 2 Peter (namely those found in Papyrus 72, Codex Sinaiticus and Codex Vaticanus) is made in order to determine how the meaning of 2 Peter is affected by differences among the three copies, especially the textual variations among them. These textual variations produce significantly different understandings of Jesus in the three copies of 2 Peter, as well as other less prominent differences in meaning.
This is an interesting attempt to read the early manuscripts and detect tendencies within their peculiar readings, especially in relation to Christology. Although it is not very well-informed about text-critical matters, utilises a method I would regard as fundamentally flawed, and misses some significant data, it does raise some interesting questions.
The article contains a number of errors and misleading statements. E.g. that Ehrman and Royse follow the same approach to variants and theological tendencies (p. 427 note 2), that the pages of P72 are half the size of those of Sinaiticus and Vaticanus, that Sinaiticus is lacking all of Genesis to 1 Chronicles, that Sinaiticus has 47 lines of text per column. He also adopts disputed positions without considering any alternatives, e.g. that the same scribe wrote Jude and 1&2 Peter in P72, that Sinaiticus and Vaticanus are from the same scriptorium. These are a little irritating, but don't impact the overall argument very much.
As for method the intention seems to be to consider what '2 Peter would have meant to the readers of each one [manuscript]. My concern will be the meaning of the manuscripts as they stand.' (p. 428). But attention is only given to variants between the manuscripts readings and the text of UBS4. It is a bit like a severe form of redaction criticism which assumes the reader moved from one "significant" variant reading to the next "significant" variant reading, and doesn't treat the text as a whole. In addition he is rather inconsistent in deciding to treat the corrected text of Sinaiticus (except when a text is dotted for omission). On this basis the corrected text is not as early as he thinks, and since he is reading both original words dotted for omission and other corrections, he would seem to be reading a text that no one in antiquity actually ever read.