Feasibility question (involves mainly I2S)

[morcilab]

I have the current audio setup:
A RPi Zero 2 acting as a USB UAC2 device that receives a sound stream via USB.
The stream is processed (using CamillaDSP) and dispatched to an I2S audio pseudo-device.
The RPi acts as an I2S clock slave, the three clocks (master clock - used by the DAC, ws clock and sd clock) are generated by an Si5351 (controlled via I2C).
The I2S stream is then directed to an I2C-controlled DAC/AMP (the discontinued MA12070p) which is connected to the speakers.
The goal is to get rid oh the RPi.
I'm trying to understand if Rust on the ESP32 (ESP32-S3, specifically) is the way to go.
From the info I gathered around the net the ESP32-S3 can act an an I2S clock slave, it's less clear whether it can act as a UAC2 device. But I also do not want to deal with C and its callback hell. I'm looking at Rust mainly because of async support (and yes, I'm familiar with the language outside the micro-controllers world).
Before wasting the next two weeks testing and gathering info, I'd like to hear the opinion of somebody more knowledgeable than I am at this point: do you believe this goal is feasible? Thanks.