cognitivedaily is reporting on an experiment, which sheds some extra light on the inference of conscious will:
Here, the subject presses the button, and the beep sounds as a result, but a random delay is added between pressing the button and the resulting beep sound. Then the subject is asked to pinpoint the moment of time, when the decision to press the button was made. It turns out that subjects on average estimate that moment as approximately 130 milliseconds before the beep sound (and that the estimate depends very weekly on the actual time when the subject pressed the button).
So this is another illustration how the inference of conscious will uses the observation of the physical event which is (apparently, and in this case, actually) caused by the subject. Wegner's experiments were also based on this, only in his case the event only seemed to be caused by the subject, while in this case the causation is real, but a random delay is inserted.
I was trying to interpret the 130 millisecond number, which seems to be pretty close to the characteristic time intervals for visual perception, or motor control signal propagation. Of course, it is difficult to say anything conclusive. But this time interval might be close to the "correct estimate" of the interval between the initiation of motor control neural circuits and the actual movement in "naturally occurring situations".