The primary advantages of being file-based is hardware independence and single-instance storage of a file referenced multiple times in the filesystem tree.
The cost of reading or writing many thousands of individual files on the local disk is negated by hardware and software-based disk caching as well as sequential reading and writing of the data.
WIM files can contain multiple disk images, which are referenced either by their numerical index or by their unique name.
[5] WIM supports three families of LZ77-based compression algorithms in ascending ratio and descending speed: XPRESS,[6] LZX, and LZMS.
It allowed Microsoft OEM partners to experiment with the imaging technology and was developed in parallel with Longhorn alpha prototypes.
Build 6.0.5384.4 added significant advantages over previous versions, like read-only and read/write folder mounting capabilities, splitting to multiple image files (SWM), a WIM filter driver and the latest compression algorithms.
Thanks to this project, Linux distributions now have their own imagex clone called wimlib-imagex, which allows mounting WIM images and managing them (read/write) like any other block-storage provider.
[15] As WIM images use somewhat common compression algorithms, they can be accessed by using file archivers like 7-Zip.