I can only guess both "40% and 80%" are not correct:
overlap is an esitmation. It can be larger or smaller than the actual, physical coupling coefficient.
For EME, I am not sure how many number of modes you used and what is the structure of the waveguide. We have an online example: https://optics.ansys.com/hc/en-us/articles/360042305354-Edge-coupler
You may need to do converge test, with sufficient cells and mode numbers. Of course, if there is no strucutre change, only one cell is sufficient.
You are navigating away from the AIS Discovery experience