Aune on the Text of Revelation

I have been impressed with Aune's text-critical work in his commentary on Revelation. It contains perhaps the most thorough and independent text-critical judgement of any commentary of the last thirty years. (Discuss?)

One general point he makes puzzles me and I wonder if anyone else can shed any light on it. It relates to the important fact that a large number of the manuscripts of the Apocalypse are found within collections of non-biblical writings.[1] In this connection Aune states:

• ‘The text of Revelation was transmitted in two quite different settings, in an ecclesiastical setting and as part of collections of miscellaneous documents. Copies of Revelation transmitted in the second way have been less subject to harmonistic and theological alterations.’ (Revelation, p. cxxxvi - my emphasis)

So my problem is that I can't see any correlation between the list of Revelation manuscripts which were transmitted in miscellaneous collections and supposedly less harmonistic alterations (many of them are fairly straightforward Byz). Nor can I see any correlation between these manuscripts and the 34 places where Aune proposes different readings compared with the NA27 text, most of which seem to reflect a different perspective on the internal evidence.[2] Nor can I see any correlation between this statement and the general perspective (shared not only by Schmid, but also Weiss, Charles, Kilpatrick and Aune) that Alexandrinus contains the best single text of Revelation.

[1] Schmid, Studien zur Geschichte des griechischen Apokalypse-Textes, II.33-35: Q, 920, 1774, 2076, 2258, 1806, 2015, 2016, 2017, 2024, 2025, 2048, 2049, 2050, 2057, 2061, 2079, 2084, 2196, 2329, 2436, 2078, 2377, 2434. To which Aune apparently adds: 2023, 2024, 2030, 2449, 2681, 2737, 2743 (Revelation, p. clviii).

[2] Aune lists 40 (pp. clix-clx), but the variant at 9.9 doesn’t exist (duplication); 17.10 is an error for 7.10; and at 18.16; 19.5, 12; 21.12 he actually agrees with the NA27 text (and simply proposes deleting square brackets); at 20.4 he also agrees with the NA27 text (and has mistakenly represented it). This leaves: 1.6; 2.15; 4.4, 7, 8; 5.6 (2), 10; 6.17; 7.10 (kra/zousi, not listed in NA27 app.); 9.6; 10.6; 11.16; 14.13, 16, 18 (2); 16.4, 6 (2) (pei=n, not listed in NA27 app.); 17.3 (2); 18.2, 3; 19.6, 7, 9, 11, 17; 20.11 (au0tou= after prows=pou, not listed in NA27 app.); 21.16, 22, 27; 22.11