I would recommend double checking them with an RCL meter, if you have one on hand.
One possibility is that during the inductance measurement, the inertia of the drive train keeps the motor locked at a different rotor position that has a different corresponding inductance.
The best practice would be to do detection without the drivetrain attached because potential inbalances in the drivetrain can cause the motor to move during detection and skew the values by the generated bemf.
Our motors probably aren’t too similar, but for what it’s worth, the TB6380 I have, measures a phase resistance of 21miliohms (power stage included, it is about 37.5miliohms line to line from the bullets), phase inductance of 15-17uH at 30khz (19uH at 1khz and low current). And flux linkage of .00475Wb. These are measurements taken using an RCL meter and a separate power stage.