I also prefer to number just the important equations. One of the advantages of numbering all equations is that the reader can binary search an equation by numbe

I also prefer to number just the important equations. One of the advantages of numbering all equations is that the reader can binary search an equation by number. I try to mitigate that loss by providing page numbers or section numbers. Providing “contrast” is something I appreciate as a reader. It is much easier to provide in person, on a whiteboard. Maybe we need papers to have references to YouTube derivations :)