This post is not specific to Seamly however it may be of interest to some.
This paper was presented at this years Siggraph (2018) Vancouver Canada.
Warning… This paper is math based
Huamin Wang. 2018. Rule-Free Sewing Pattern Adjustment with Precision and Efficiency.
ACM Transactions on Graphics (SIGGRAPH), vol. 37, no. 4, pp. 53:1–53:13. Link below
I found the paper interesting, though it is almost an opposite approach to Seamly… the grading of multiple sizes is based on an “average” size and shape. In my experience there is nearly infinite variation in size and shape in real people and the difference is not expressed by a continuous mathematical function
One technique I have used over the years is to start with multisize printed patterns and use freehand drawing or a french curve to join sections where a particular body needs multiple sizes to fit. Using Seamly2D as I tool, I find it easy to “morph” a pattern to fit myself or another who is not shaped like the “average” pattern size