Today we made further modifications to the ISS box, and implemented it in the setup again.

Modifications:

We found that the suppression of the intensity noise was not optimum last time, due to positive feedback. The problem arose from an additional portion of the box contributing to the transfer function, other than the filter stages. There was a high pass filter preceding the vga on the second page of the schematic (U8). We removed the op amp for this stage, as well as R38, and added a wire from TP9 to TP11. After spending some time mapping out the circuit, I found that TP11 goes directly to the vga input. 

Transfer Function of Box:

After modification, the transfer function for the box in the state we are currently using it is:

It's interesting that the phase displays positive feedback, because there is a flip sign stage after the filter stages, which when set opposite to this does not enter positive feedback at all. However, suppression is only achieved with the phase in this orientation.

Measured RIN and Calculated Frequency Noise:

The following plot shows the suppressed RIN vs free RIN:

The corresponding suppressed RIN Noise vs free RIN Noise plotted with the current noise budget is:

There is about a factor of 4 suppression here.

I might think about changing the corner frequency to get more suppression at lower frequencies.