While looking for possible enciphering schemes used by Z340 (in addition to assumed homophonic substitution) I came across this interesting "connection". British Army had an encryption system that used a matrix of 17 rows by 12 columns:
https://en.wikipedia.org/wiki/Slidex
http://www.cryptomuseum.com/crypto/slidex/index.htm

Well, Z408 is exactly 2 sheets of 17 rows by 12 columns. And all other known Zodiac ciphers are 17 columns wide. I’m not quite sure what to think of this "coincidence". There is no logical reason (that I know of) why Zodiac used 17 characters per line, other than it’s a prime number. And in the Slidex encryption system, the 17×12 matrix was only used for encoding the plaintext. The final ciphertext that came out after encoding didn’t have any predefined dimensions to it. Slidex was extensively used by the British in World War 2, so it not impossible that Zodiac knew, or even used it. Did he have some connection to British Army? Did he use Slidex encryption in the past and had the 17×12 matrix dimensions stuck in his mind for good?

It could also be that, as I’ve seen suggested elsewhere, 17 is simply the number of columns you get if you divide a standard letter-sized paper (8.5" wide) into 1/2-inch columns. No logic, just plain old happened to fit on a piece of paper Zodiac had on his table at the time. But it got me thinking of something else that might be interesting. What if Z408 is supposed to be somehow involved in the process of decoding Z340 (and all other Z ciphers)? Can it be that decoded Z408 is supposed to be used as a Slidex matrix to decode Z340? I.e. symbols in Z340 refer to rows and columns in Z408’s decoded plaintext somehow? Or, can it be the "decoder book", similar to how the second Beale cipher used Declaration of Independence perhaps? We just need to establish the order of symbols in the ciphertext, to be translated to position in the Z408. But I can’t think of any logical ordering scheme for circles and squares, or the +, or the zodiac symbol itself for that matter. Any other ideas of how symbols in Z340 can be referencing Z408?

After all, Z408 was just a bunch of ramblings without any real meaning. He didn’t convey any important information, for example, no details of his crimes, no hints to his identity, pretty much just a bunch of random thoughts. But it had enough coherency to verify that the decryption system was correct and true. So if he wanted to create a "crack-proof" system with increasing degree of complexity, it wouldn’t be illogical if later ciphers relied on previous ones being decoded first. The earlier ciphers wouldn’t contain any meaningful info, but later, more complicated ones, would. He just probably didn’t anticipate that the first cipher was cracked so quickly, so he might’ve cranked up the complexity with Z340 just a tad too much. Possibly even so much that it can’t be decoded even if the key and the encryption scheme is known (i.e. the encryption process he devised is inherently ambiguous and can’t be reversed).