I think the idea is the same. But we tried displaying the input signal using oscilloscope, and the 3.3V output of the microcontroller doesn't seem to be degraded. What really happens is that, each time another servo moves, the ground (0V) level of the input signal becomes too noisy. That's what I think causes jitters, because the servo seems to interpret that noise as a control signal. Also, what we do is that we placed the buffer just after the microcontroller, unlike the buffer-booster placed midway at the servo wires. Isn't it that buffer only reduces loading effect? Does it also blocks noises from coming in?
Anyway, the buffer works. We hope the jitters won't come back, because the presentation will be 2weeks from now.