One reason why it is very helpful to name things during rigging in an easy to understand way is to help make changes faster.
For example suppose we've got an arm rig with lots of set driven keys and joints etc, but we are just interested in tweaking one of the sdk's. Without naming would require to search in hypergraph connections then find the distance node if that was used etc. With naming we can type in script editor "select -r l_arm_distanceBetween" and graph its connections then quickly find the animation curve we want to tweak.
Another reason its helpful to name things is we can reuse work done on one half of rig. Example mirroring set driven keys, blendshapes etc.
Hope this was helpful,