A well known observation is that the 340 has 9 rows without repeats, these are rows 1, 2, 3, 7, 11, 12, 13, 15 and 20.

Is there anything special about this observation?

01. HER>pl^VPk|1LTG2d <---
02. Np+B(#O%DWY.<*Kf) <---
03. By:cM+UZGW()L#zHJ <---
04. Spp7^l8*V3pO++RK2
05. _9M+ztjd|5FP+&4k/
06. p8R^FlO-*dCkF>2D(
07. #5+Kq%;2UcXGV.zL| <---
08. (G2Jfj#O+_NYz+@L9
09. d<M+b+ZR2FBcyA64K
10. -zlUV+^J+Op7<FBy-
11. U+R/5tE|DYBpbTMKO <---
12. 2<clRJ|*5T4M.+&BF <---
13. z69Sy#+N|5FBc(;8R <---
14. lGFN^f524b.cV4t++
15. yBX1*:49CE>VUZ5-+ <---
16. |c.3zBK(Op^.fMqG2
17. RcT+L16C<+FlWB|)L
18. ++)WCzWcPOSHT/()p
19. |FkdW<7tB_YOB*-Cc
20. >MDHNpkSzZO8A|K;+ <---

Significance wise, compared versus randomizations it is a 7.01 sigma observation which is significant. And compared versus a plaintext + sequential homophonic substitution with 26% cycle randomization hypothesis it is a 1.19 sigma observation. It correlates with sequential homophonic substitution because the sigma went down.

Though the comparison is not fair since we have not included any offsets. Why is that? If you offset the entire cipher by 1 to the right you get a different number of rows that have no repeats. This has to do with breakpoints and thus is prone to outliers. The next table shows the number of rows that have no repeats for every unique offset:

Row length 17: 9, 7, 5, 4, 9, 8, 9, 8, 8, 9, 7, 6, 4, 3, 3, 6, 7

Notice that there are 3 offsets which have a 9 count, and that there are 2 with a 3 count. The average of these numbers is 6.58, and this average is less prone to outliers. Significane wise, compared versus a plaintext + sequential homophonic substitution with 26% cycle randomization hypothesis it is a 0.02 sigma observation. In other words, per the encoding hypothesis, the 340 behaves exactly as expected. And the 340 likely just happens to occupy a favourable offset.

An additional observation about these rows is the 1, 2, 3 and 11, 12, 13 pattern. I guess that the speculation about this pattern varies. Some may think of it as a possible encoding restart, of a start of different encoding which so far has not been detected by any means (and we can). My guess is that many random configurations of 9 rows without no repeats in one way or another can be seen to occupy a pattern of some sort.