> This is THE book about code breaking. Very concise, very inclusive, and easy to read. Good references for those who would make codes, too, like Kryptos.
That sounds like a strong hint that the book comes close to discussing the kind of code used in the remaining unsolved section, though if so I suspect Dunin would have realized it in the writing process. (Others have already attempted and failed to identify the cipher through the process of elimination.)
For a good casual book on the subject, consider Simon Singh's "The Code Book," or if you want something more comprehensive, David Khan's "The Codebreakers" is extensive (though it was first printed before public key cryptography was widely known and completely reset the field).
If you're interested a good story about a one-man struggle to get British secret agents in WWII to start using reasonably secure codes (especially unbreakable one-time-pads), I recommend "Between Silk and Cyanide" by Leo Marks.
A stronger version of the British double-transposition code that Leo Marks hated is the Soviet VIC cipher, which only has about 38 bit key strength, but is surprisingly strong for a pencil-and-paper method. For pencil-and-paper plus a deck of playing cards, there's Danial Shiu's improvements on Bruce Schneier's Solitaire.