Imagine your doing analysis of a map plotting efficiently as you can a new road.
You might want to asses the relationships relative to travel and requirements trying to taking into account local concerns.
So you square grid map the map and do your best relative to this square referencing.
The way in which the requirements are meet isn't lightly to be properly relativistically representative of every measurement required for all the more refined and fiddly bits of getting the road right.
Instead of breaking the map up into a square grid you could take multiple sized varied precision circular measurements of the map information. This more free more optimal higher relativistic circle property organisation approach allows for better management in terms of avoiding redundent calculation whist being able to handle better complexity.
If you had a plot of points and wanted fit a curved geometry range to the data then this type of circular measurement and generation of curves to fit the structure. More efficient fits to advanced data can help computers to do more optimal AI tasks.
Spherical geometric reduction based
rendering could allow for many more tricks in GPU physics and graphics.
Yes a circle is a simple part of the mathematical rudiment but we should not assume we even use rudimentary math optimally unless there's proof relative to a particular math op.