So the only thing I did differently on this build is wire the OLED to battery positive instead of 3.3v pad.
I didn’t see how that would matter, since the OLED is rated up to 5V, but there must be something else at play (that I still don’t understand).
Basically, having the OLED VCC pad direct to battery positive, with the other 3 lines (GND, SCL, SDA) hooked to the board caused freaky stuff.
The kicker (and clue) was the RCP port.
Somehow, I was getting battery voltage across the 2 negative legs.
The other problem was when booting off USB only, I got nothing at all in the serial monitor window, pulling the kill key didn’t make a difference at that point.
Conversely, If I started with battery running things, then plugged in USB, all was fine…UNTIL I inserted the kill key, at which point I got Motion Timeout messages repeating until I pulled the USB.
The only way things worked correctly was booting from battery power (pull kill key).
There must be something inherent in the display that somehow passes voltage or backfeeds or SOMEthing.
Use the 3.3v pad on the Proffieboard to power an OLED.
Below is my original post that I was composing as I was working through it.
Eventually , I read back and (as is typical with me) realized I had answered my own question with the very first statement.