Evaluate usage of R User Directories for storing larger datasets

[brownag]

Consider storing data in the standard local directory returned by tools::R_user_dir(package = "soilDB", which = "cache") i.e. "C:/Users/<User.Name>/AppData/Roaming/R/data/R/soilDB"

Would probably be useful to provide a couple functions to streamline access, versioning, and cleanup of the cache.

For starters, we could cache the MLRA polygons that are returned by fetchSDA_spatial(geom.src = "mlrapolygon") downloaded from https://www.nrcs.usda.gov/sites/default/files/2022-10/MLRA_52_2022.zip

This would be helpful for providing more realistic testing datasets, and also separating large or complex datasets from the package installation itself--making the installation smaller and less prone to issues.