Spatial (resolution-aware and context-aware) caching

Spatial, and likely temporal, caching needs to with based on storing cached data at different zoom levels, areas in space (forming a patchwork quilt of cache hit areas within surroundings of cache miss areas), and different times in the animation timeline. All this will be needed to make Cache nodes relation-aware to work with footprint data.

Start by mapping the existing Cache nodes and relation-aware footprint handling in the graphics engine. Define how cached data should vary across zoom levels, spatial areas, and animation times, then document tests that demonstrate patchwork cache hits around misses and correct invalidation across timeline changes.

Written by the indexing model from the issue text.