From you request, I ran Bruco on those times. The results are as follows,

bad time (12:20 UTC) : https://ldas-jobs.ligo-wa.caltech.edu/~youngmin/BruCo/PRE-ER10/H1/Oct24/H1-1161346817-600/

good reference(11:08 UTC): https://ldas-jobs.ligo-wa.caltech.edu/~youngmin/BruCo/PRE-ER10/H1/Oct24/H1-1161342497-600/

These could give you a hint for range drop.

Here is a plot of thge auxiliary loops, again comparing good vs bad.

Note the two broad noise humps around 11.8Hz and 23.6Hz. They both increased at the bad time compared to the good time.

Interesingly, the peaks showing up in the DARM spectrum are the 4th 5th, etc. to 12th harmonic of that 11.8-ish Hz.

It all smells to me like some form of scatter in the input chain.

The IMs do not change their motion between Stefan's good time (11:08 UTC) and bad time (12:20 UTC).  But, the RMs, particularly RM2, see elevated motion, almost a factor of 2 more motion between 8Hz - 15Hz.

First screenshot is IMs, second is RMs.  In both, the references are the good 11:08 time, and the currents are the bad 12:20 time.

Stefan and TeamSEI are looking at the HEPI and ISI motion in the input HAMs right now.

EDIT:  As one would expect, the REFL diodes see an increase in jitter at these same frequencies, predominantly in pitch.  See 3rd attachment.

I quickly grabbed a time during O1 when this type of noise was happening, and it also corresponds to elevated motion around 6 Hz in RM1 and RM2. Attached are a spectrogram of DARM, and the pitch and yaw of RM2 at the time compared to a reference.

There is a vertical mode of the RMs at 6.1 Hz (that's the LLO value, couldn't find it for LHO). Maybe those are bouncing more, and twice that is what's showing up in PRCL?

There should not be any ~6 Hz mode from the RM suspensions (HSTS or HLTS), so I am puzzled what this is. 
For a list of expected resonant frequencies for HSTS and HLTS see links from this page

https://awiki.ligo-wa.caltech.edu/aLIGO/Resonances

@Norna: the RMs, for "REFL Mirrors" are HAM Tip-Tilt Suspensions, or HTTS (see, e.g. G1200071). These, indeed, have been modeled to have their highest (and only) vertical mode at 6.1 Hz (see HTTS Model on the aWiki).

I can confirm there is no data committed to the SUS repo on the measured vertical mode frequencies of these not-officially-SUS-group suspensions at H1. Apologies! Remember, these suspensions don't have transverse / vertical / roll sensors or actuators, so one have to rely on dirt coupling showing up in the ASDs of the longitudinal / pitch / yaw sensors. 

We'll grab some free-swinging ASDs during tomorrow's maintenance period.

Stefan has had Hugh and I looking SEI coupling to PRCL over this period, and so far I haven't found anything, but HAM1 HEPI is coherent with the RM damp channels and RM2 shows some coherence to CAL_DELTAL, around 10hz.  Attached plot shows coiherence from RM2_P to HEPI Z L4Cs (blue), RM2_P to CAL_PRCL (brown), and RM2_P to CAL_DELTAL (pink). The HAM1_Z to PRCL is similar to the RM2_P to CAL_PRCL, so I didn't include it. HAM1 X and RY showed less coherence, and X was at lower frequency. There are some things we can do to improve the HAM1 motion if it's deemed necessary, like increasing the gain on the Z isolation loops, but there's not a lot of extra margin there.

Here are ASDs of the HAM3 HEPI L4Cs (~in-line dofs: RY RZ & X) and the CAL-CS_PRCL_DQ.  The HAM2 and HAM1 HEPI channels would be assessed the same way:  The increase in motion seen on the HAM HEPIs is much broader than that seen on the PRC signal.  Also, none of these inertial sensor channels see any broadband coherence with the PRC, example also attached.

Freee swing PSD of RMs and OM are in alog 30852.