disclosure: I am a contributor to LCFS. One of the things I thought was interesting about this update is how LCFS leverages its layered architecture to solve the problem of scalability faced by other storage drivers. Irrespective of the size of docker container a docker commit would always take the same amount of time.
SSD was used as the backend device for the tests.
If a file is used for LCFS backend, the performance will be limited by the underlying file system in which the file is hosted. LCFS is capable of working directly on top of device and can queue I/Os efficiently and avoid duplicate caching. One of the reasons for developing LCFS as a native file system was to avoid some of the disadvantages of a merged solution.