From the convex picture, it looks like the knife isn’t exactly quite perpendicular, which would account for the wide reflection on the “right” side and almost nothing on the left.
Also, even if it is convexed, the edge of the edge should still split the beam, even if it spreads it out as in right side of the convex. The highest reading should be the edge angle, while the convexing of the bevel should produce the spread of lower angles. (Don’t forget, while the edge may approach zero degrees, the actual edge angle increases on convex grinds).
I’d play with the knife position on the Falkniven a little more.
