Something generally in the concept of Tanaka's illuminated contours ( https://mountaincartography.icaci.or.../kennelly1.pdf has some information ), except with shading direction and line width based on local slope angle and steepness rather than illumination from a fictitious light? I recommend trying a few maps to show this idea, maybe write up a paper with those maps plus the technique and submit it somewhere. Lots of conferences are looking for submissions, even (sometimes especially) for novelty elements.

I suspect that one of the reasons for apparent general disinterest from publishers is for practical reasons. You'd need at least two tightly-aligned process colors on the map or go with what amounts to a raster image and then halftoning or other process. Contour maps are cheap to print because they don't use a lot of ink or require a lot of colors. They are also well-understood by their user base and a new, potentially easy-to-misunderstand implementation might not be much of a seller. By the time that you're using multiple colors, a lot of publishers are more likely to go for existing hillshading variants of one form or another. It's not a high-margin business for the most part and they tend to be pretty conservative in what they publish.

Older (19th century or so) maps used something called hachures to show local slope and direction, with some attribute of the individual hachures (usually thickness, sometimes length) coding slope. You can arrange the hachures along contour lines or encode density of hachures to show various attributes. Because they generally where's precisely defined, they largely fell out of favor with contour maps as their replacement. Contour maps have the nice property that the steepest up/down local slope is always at right angles to the contour.