Posting this because hopefully others—with sharper minds than me—can make some use of these little mistake maps I made with doranchak’s lovely cipher explorer. You know it is good because it is still proving useful even after the solve

Please ignore the parts with only the white background in the ciphertext image. I forgot that I had selected them to see more clearly the locations of the ‘outlier’ parts "life is" and "death" in the transposed ciphertext—and I didn’t even select the right location for ‘life is’

White text/black background = likely incorrect ciphertext symbol selection (intentional or unintentional—but unlikely to be errors in plaintext spelling).
Red text/turquoise background = double use of a ciphertext symbols for plaintext A when plaintext should be U.
Turquoise text/red background = use of ciphertext symbols for plaintext V when plaintext should be K (unlike plaintext U, however, these are the only uses of plaintext K – so perhaps Zodiac did not want to assign them a unique ciphertext symbol because of their infrequency or perhaps just laziness or poor planning).
Yellow text/green background = symbol one spot earlier/to the left of expected position because of missed symbol in its position.
Green text/yellow background = symbol which should have been in that position.
Dark green text/peach background = all symbols which were affected by the missed symbol shift.
Lilac text/dusty blue background = good old fashioned Zodiac spelling errors, almost certainly just in the plaintext (i.e. the c in paradice).

There are probably errors I’ve missed, maybe some more spelling errors. You get so used to some Zodiac misspellings of words that you stop seeing them as misspelled, lol!

I am particularly interested in what the yellow/green/peach section tells us about how the cipher was constructed and in what steps.

For example, I like doranchak and others’ suggestion of the 9 x 17 columnar transposition followed by the split triangular reversal thingamabob (boy, was it jarlve’s idea originally/before the solve? I don’t know sorry). There is an issue, though, when we consider how he has accidentally missed a symbol and this has affected the entire rest of the row (in the dark green/peach text) rather than simply the triangle in which the missed symbol occurs. This is reinforced when we consider that he was then able to insert the missing symbol (green text/yellow background) at the end of the entire row, not at the end of the split triangle row or at the end of the entire left side triangle, for example.

I can think of two explanations for this. The first is that the 9 x 17 columnar transposition followed by the split triangular reversal thingamabob was not the method Zodiac used to encipher the text. Maybe there is an encryption method method which would’ve made it more likely to mess up in the particular yellow/green and green/yellow section kind of way. The second explanation is probably the right one, has some other supporting evidence, and honestly makes it no easier to be certain about the Zodiac’s encryption method for the 340. The second explanation is basically just that the zodiac enciphered the 340 in a rough draft initially and copied it out into the polished final draft, but missed the symbol at the yellow text/green background spot, continued copying it out neatly and only realised his mistake when he got to the end of the line and there was a gap there, looked back over his rough draft and realised which symbol he had missed and just chucked it on at the end. I think the neatness of the 340, combined with the sloppiness of other error corrections (like the crossing out with the backwards k) makes it pretty likely this is what happened.

But the point is that this error provides us with data points about the process used to construct the 340. What else is there that we can learn from the errors? After all, "we learn wisdom from failure much more than from success." The outstanding efforts of all involved in decrypting the 340 over these last 51 years are certainly testament to that.