I do have a theory..
Look at the pictures below:
- Blue line is stator output voltage at low rpm.
- Red line is stator output voltage at a higher rpm.
- Black line is the battery nominal voltage.
- Green line is the charging voltage R/R is trying to achieve.
The first picture shows one positive half wave of the stator output voltage.
The second picture is a detail of the first one, where the voltage rises beween the battery voltage and the charge voltage.
While the voltage is below the battery voltage, there is no output from the R/R and a voltmeter will show the battery voltage. As soon as the voltage rises above the battery nominal voltage, the battery starts to receive charge current. But when the voltage goes above the R/R set voltage, it cuts off the R/R output. I have colored the triangles where charging occurs. As you can see, at low rpm the triangle area is larger than at higher rpm. Since a voltmeter shows the root-mean-square (effective) value of the voltage, it will actually show a higher reading at lower rpm. I used the same timescale for both cases (lazy me), but actually the dirrefence is even more when the blue curve is stretched. On the other hand, at higher rpm those red charge pulses occur more often, which compensates.
In practise the stator output voltage is not as clean as in the pictures. As soon as it gets above the battery voltage, it is the stator which supplies all the power + the charging current. This tends to restrict the rise of the output voltage. But with a battery which has seen it's best days, less than perfect connections to the battery and a light load, you could see a lower voltage at higher rpm.
Did it make any sense to you? Could explain what niterider is seeing.