The shortest honest answer: pick the highest setting the converter offers, and know that the original video caps everything. YouTube already compresses its audio, so no converter setting can restore what the upload lost. What follows is how to pick a setting that actually fits what you are converting.
Every video you paste in arrives with audio that has already been compressed once, by YouTube, in a lossy format. Think of the converter as making a copy of that copy. A bigger output setting preserves more of what survived, but it cannot invent detail that was discarded upstream. This is why "high quality" here means "least additional damage," not "better than the video." The 320kbps question, answered honestly.
Bitrate choice is really a file-size versus fidelity trade, and the right answer depends on what is in the audio:
The point of diminishing returns is set by the source upload, not by your setting. If the original video had thin, tinny, or noisy audio, a 320kbps output will just be a faithful copy of thin, tinny, or noisy audio. The setting controls how much additional loss happens on top of what YouTube already took. Once the setting is high enough that its own damage is negligible, pushing it further changes nothing you can hear.
For the full picture, read the bitrate comparison and what the 320kbps setting really does. The pattern is consistent: choose by content, not by habit.
The ceiling is the source audio. Use the 320kbps setting for music; it preserves the most of what the source offers.
For music, yes: higher settings lose less detail in the re-encode. For speech, 128kbps is usually enough.
Audio science: MP3 file size per minute, Opus vs AAC vs MP3, can you hear the difference. Diagnostics: why converted MP3 sounds bad. Playlists: YouTube playlist to MP3.