Hi ahanford,
To get an idea about the differences quickly, please use the link below and scroll down a little bit until you see the overview (attached as a small preview).
https://help.maxon.net/r23/en-us/Default.htm#html/MMATERIAL-ID_MATERIALGROUP_REFLECTION.html#REFLECTION_LAYER_DATA
Focus on the 25% and 50% values first. Here, have an eye on the two dots in each image in the lower right corner. Then scroll a little bit more down tow the image with the four heads. The differences might be subtle, but that is the point, to get the exact character of the material you need.
The next step would be to go through this series to understand how to mix all Layers into one material.
https://www.cineversity.com/vidplaylist/the_reflectance_channel/reflectance_part_01
In this series, it is mentioned that one can get values above 100%, which requires to render into a format that can differentiate values above 100%. Integer based images can’t do that. If the format is limited to 0-100%, the misguided idea of a black or white point is often given. Which is not natural nor physically correct at all. When every bright part is clipped at 100%, it is falsely labeled as a white point. Any differentiation later on in color correction is then cut off. I mentioned this here, as an evaluation of the material is then limited or even disabled. This means, in return, the idea of what a material can do can’t be understood to the fullest. Confusion about the materials is the result. If you render a material, see its quality in a scene (after all, the use without the context of reflective material is kind of limited). Use the Filter options in the Picture-Viewer to explore what the screen can’t show you, except you use an HDRI monitor.
I hope that helps.
All the best
