Actually, I believe that the motor needed a lower voltage (also as the op requests), feeding full voltage PWM does not reduce the voltage. Whether the $40 critter control's full voltage pulses would have damaged the motor is unknown.
In Z scale, we have motors that often have a maximum of 9 to 10 volts, and people often use a DCC supply of 13 volts, but because the motor speed is controllable, they forget that they are still sending 13v pulses to a 9v motor. It does not normally end well.
Again, we are making some assumptions about the motor that cannot be proven, yet make sense from the observed information.
Greg