The reason the graph appears to be identical for torque and backEMF is because for the ideal case \$K_e = K_t\$ & it is an over-simplified video, plus some external controller deals with the stator energising polarity.
In practice they are different as \$K_e\$ is defined as the open terminal voltage while \$K_t\$ is usually taken at maximum design current (and thus has some aspects of stator saturation)
To generate reasonable torque the controller MUST align the stator voltage to the rotor backEMF otherwise you end up with a phase-advance/retard which is essentially field weakening.
So the controller MUST energies the stator in-phase with the rotor (its a sync machine at the end of the day so the frequency must equally be equal).
Why are they the same profile? as stated \$K_e \approx K_t\$ and equally the controller firing the stator must provide a similar voltage profile to the BackEMF to maximise electrical --> mechanical conversion
The y-axis scaling however will be different.