Abstract:

We injected sine excitation at 6.125Hz to increase the RMS of OMC transmission.
Though RMS was increased by a factor of 5.7, no significant excess can be seen around 60-100Hz, so low frequency RMS doesn't seem problematic.
I also estimated the siderobe due to the low frequency spectrum and it seems negligible as well.

Detail:

We measured the spectrum of calibrated DARM displacement and OMC trans power by injecting the sine signals from DARM error point.
Excitation frequency was 6.125 and amplitudes are 10e-6, 20e-6, 30e-6, 40e-6, and 50e-6 cnts.
Figure 1 shows the measured spectra (excitation amplitude is written in the legend and blue lines are the spectrum without excitation for the reference).
Even though the RMS was increased by a factor of 5.7, no significant excess can be seen around 60-100Hz.
So, DARM non-linear coupling due to low-frequency RMS doesn't seem to contaminate the sensitivity around 60-100Hz.

In addition, we evaluated the siderobe of low frequency spectrum.
When we injected the line at 6.125Hz, 50.875Hz peak was appeared in OMC trans spectrum, which generated due to non-linear coupling between 44.75Hz peak and 6.125Hz peak (see fig2).
According to the peak height at 50.875Hz, sideband peak height was propotional to the peak height at 6.125Hz, so we can evaluate the siderobe of 44.75Hz peak from the OMC trans spectrum.
Figure 3 shows the estimated siderobe generated due to 44.75Hz.

Though the noise floor is low enough, some peaks are very close to the current sensitivity.
So, damping/supressing the peaks aroun 20Hz and 40Hz seems effective to mitigate the sensitivity degradation when suspension resonances are kicked.
In addition, each peak should have a similar effect, so the summation of siderobe would be contaminate the sensitivity even during steady state.

So, it would be worth trying to damp/supress the resonant peaks.