access to much more memory than available as RAM, potentially gigabytes. (Although for this to scale without using lots of RAM, the sector size may need to be increased (into clusters) to reduce mapping table size.)
persistence built-in
familiar API - String/Vector/Map etc.. are well-documented.
cons:
degraded performance - most of these devices are not as fast as RAM, plus the overhead of delegating to a backing store
See https://community.spark.io/t/flashee-eeprom-library-eeprom-storage-in-external-flash/5241/77