I'm working on what may be best described as a puzzle cube. Each face will have various inputs, each with corresponding WS2812b pixels for target value and current value, and the player will turn the potentiometer or whatever so the target/current pixels have the same hue. Easy.

I would like to add an audio component to this, which is where I'm stuck. I am working with a T4.1 and have the matching audio shield, but I'm not sure how to proceed.

This is what I'm aiming for:
-the puzzle cube will be able to play a small number of audio files from a micro SD card (think 'success' or 'failure' notification sounds), to come out small speakers in the cube or headphones
-the User is able to play their own selected music over Bluetooth, to also come out of the same embedded speakers/HP
-the cube will be able to analyze the audio coming over BT for basic FFT and BPM purposes to lightly influence gameplay
-while handling the above functions, the Teensy will also need to be reading varying combinations of several different inputs (A/D) and sensors (I2C), and occasionally updating pixels as required

My questions:
1. What, if any, is the difference between storing the game sounds on the micro SD on the Teensy vs. micro SD on the audio shield? Pros/cons?

2. What is the best way to route the mono audio signal from the Bluetooth receiver to the output headphone jack on the cube? (This is a separate jack from the Audio Shield, and will be normalled to the speakers.)

2.a) If I route BT-> Audio Shield Line In, will this necessitate the Teensy constantly processing the incoming audio in order to output the music at Line/HP Out, or is there a passthrough not shown in the schematic?

2.b) Does it make more sense to route the BT signal both to Audio Shield Line In AND directly to the output HP jack/speakers, and have the separate output jack also connected to Line/HP Out on the shield? (So Audio Shield can still process the signal as needed, but processing is not required for music to get to the player, while still allowing the Teensy to play game sound effects. Would this require op amp buffering/mixing, or could it be done passively with resistors?)

3. Would it be preferable to connect the output 3.5mm jack / normalled 8 Ohm speakers to the Line Out or HP Out of the audio shield? (Assuming the speakers will almost always be used over headphones, but with the possibility that at some point someone might plug the output of this cube into something expecting a line level input signal.)

If the sum of all these tasks is too much, I would prefer to scrap the BT audio input / FFT piece and keep the ability to play game-related sound effects.

I greatly appreciate any answers to these questions, thoughts on the project, or warnings of any potential 'gotchas' that might arise while trying to accomplish these goals.

(And if there's an M that I need to RTF, I will gladly do so- I just have either not yet found, or not understood well enough from reading, the answers to these questions.)