I do not know the system well enuf to answer or argue any details of how the PCM decides when to enter or exit Open/Closed loop.
O2's have nothing to do with Open Loop or Closed loop.
It's time based, temp based and throttle position based, and time since last start/soak time.
People put too much stock in what the O2's do, which isn't much really. They are an after the fact sensor only, second order sensor.
This post is meant to help me and others learn; not to argue with you; I am a fan of your work here, thank you.
I would bet that disconnecting the upstream O2 sensor would knock the fuel system out of Closed loop and into Open Loop... if so, the O2 sensor DOES have something to do with it.
Edit: O2 sensor is most definitely a primary sensor while in Closed Loop -and- if defective, would push system into Open Loop. Even the scan tool software reports "Fuel Status: Closed Loop (using O2)". Previously shown above in blue font.
time based: only thing I know is my car waits 10 seconds (while PCM is checking other sensors and heating O2 sensors right?) before entering Closed Loop.
temp based: my car entered into Closed Loop on a cold engine/45 F, if low Temps (ECT/MAF Temp adjusted?) prevent Closed Loop, it is something lower than 45 F. This goes back to the OP's question; and nothing I have seen so far proves that Open/Closed Loop is toggled because of ECT.
TPS based: I guessed this earlier, but have not tried WOT yet to see if my (cheapo) scan tool would detect a moment/switch into Open Loop. Instead of Visual, maybe I could datalog this PID (Fuel System Status) in a future test.