Hi Nihil.
One argument would be that Wwise converts the files into a Vorbis format optimised for runtime performance.
Take a look at this: https://www.audiokinetic.com/en/library/edge/?source=Help&id=vorbis_encoder_parameters
That said, by only using uncompressed audio files (PCM), you can easily re-convert into a different format should you choose to do so later in production. For instance, you may realise your soundscape is using too much CPU or battery life on mobile, and so you change some sounds into PCM to reduce CPU used on uncompressing the formats in runtime, which is needed for the sound to play.