Nope, I never read anything that way.
The software of course is ultimately what decides when to limit regen. It can do this based on any factors it wants to. It could do it randomly, it could do it based on underlying hidden actual SoC, it could do it on displayed SoC.
Without testing we'd only be guessing.
a) People have seen no regen at 100% on non-locked cars.
b) People have seen regen at 100% on locked cars.
The conlcusion to draw here is that regen isn't implemented based on the displayed % then, but some other hidden variable ... like the underlying actual SoC which is close to "real" 100% on non-locked cars, but not close to real 100% on the locked cars.
If (b) showed no regen as well, we wouldn't learn much, because it could be limiting regen due to a policy to enforce no regen at "displayed 100%" or because of no regen at near "true 100%" ... but since we have data points from people who DO see regen in scenario (b) for locked cars .... well then, we're where we are at now.