I guess there must be some theories in crystallography that may benefit from this kind of knowledge. These days a lot is being brute-forced though, so intuitive geometric arguments are getting less popular.

I suddenly also remember the video @Scroogey had linked where error-correcting codes have some link with the problem of optimally packing circles. At the time, I only watched the first one of the linked videos in #852443, I should watch the second one too.

Indeed, for logistics, I think they won't care about a 0.01% packing improvement ;)