The enhancement from (formerly Project Shasta) Adobe Podcast is AMAZING. It transforms (most) bad audio into professional sounding audio and professional audio into beautifully-curated EQ'd and Compressed, pristine audio. If you haven't tried it first-hand, you MUST give it a shot. It could make you a believer.
The good: For single-channel, vocal driven audio, there is literally nothing that sounds better to me.
The bad: 1. I have to export a Raw Waveform and submit it to the website. 2. I have to wait for the file to process. 3. If there are any anomalies that created unwanted compression/limiting/EQ, I have to keyframe the audio's gain in those parts and rinse and repeat until I get the result I want. 4. There are NO parameters to fiddle with, but the end result is usually 99% perfection, I kind of don't care: it is quickly becoming part of my editing workflow on every project.
The hope: Whether the process could live locally on Premiere or whether it could live through an integration (maybe like the Auto-transcribe function), I would absolutely LOVE the ability to let the AI of Adobe Enhance process do its magic with less steps - maybe even something along the lines of the Remix feature (which is also an amazing feature).
Anyone who has had first-hand experience with Adobe Podcast Enhancement will tell you: this is an AI goldmine for the novice and/or the rushed/impatient editor. Integrating it better into Premiere and Audition and even After Effects would absolutely push it further ahead of the pack of competing NLE's.