The bigram peak at a distance of 15 is with the cipher mirrored. Possible outlier, possibly related to the "+" symbol, haven’t checked.

My cipher manifesto: each cipher is a closed system, no information can escape from it. The information always goes somewhere. With information I mean any kind of statistics that can be derived from the cipher. Let’s take bigrams as an example. Sometimes when I create a cipher I notice that the relative bigram counts for the horizontal direction are kinda low, but then notice that bigrams at a distance of 2, 3, 4 are a bit higher than usual. This also happens when applying transposition in the mix and whatnot, the information always goes somewhere, and somehow can be retrieved.

Ofcourse it is difficult because you will also encounter outliers, and the 340 has some interesting features which really make it much harder to analyze.