I honestly can't remember the reason we never implemented reuse of audio cache files. Probably something related to cache management and preventing misassociating cache files with source files. (E.g. you load audio from file A, which then gets cached with key X, then later you load audio from file B, but somehow B receives key X too and you then get audio from A instead of the intended B.)
What Aegisub does support is direct reading from uncompressed (PCM) WAV files. If you decode the audio track you need to PCM WAV and load that into Aegisub, it will open instantly without any caching. The decoding could also be semi-automated with a batch file or similar.
If you see audio cache files hanging around after closing the audio, that's a bug. I haven't heard about that before.
Do not PM me asking for help with Aegisub, Automation, karaoke or anything like that, I won't reply.
Ask your questions in the public forums. Use the Search function, somebody probably asked the same question before.
All rights reserved to call you insults if you ask poorly thought out or redundant questions.
Experimenting is free and is a great way to learn things.