Radically different baking settings, same dataset, interesting results.
The model I think is fairly self explanatory.
Works well at a strength of .8-1.
Previews made with novaFurryXL_illustriousV5.0