General loop signal calibration trick for 
turning "CTRL" or "OUT" signal into an 
absolute measure of the equivalent out-of-loop
"INP" input signal

               ERR               OUT
                ^                 ^
                |    +-----+      |   
INP --->( + )---+----|  K  |------|
          |          +-----+      |
          |                       |
          |                       |
          |          +-----+      |
          +----------|  P  |------+
                     +-----+
                     
OUT = K * ERR
            1
    = K * ----- * INP
          1 + G
      
      G = P*K
      => K = G/P
    
       1      G
    = --- * ----- * INP
       P    1 + G 
       
      G/(1+G) ~ 1  (where G >> 1)
      
       1   
OUT ~ --- * INP
       P    
       
INP ~ P * OUT   (where G >> 1)

==>> PS = (P_PS) * ISO_OUT   (where G >> 1),
since ISO loops are DC coupled, then over hours-or-days long trends, G >> 1.

/ligo/svncommon/SeiSVN/seismic/BSC-ISI/H1/ETMY/Data/Figures/Transfer_Functions/Simulations/Damped/
H1_ISI_ETMY_TF_Damped_C2C_from_ST1_ACT_to_ST1_CPS_2014_04_01.fig
H1_ISI_ETMY_TF_Damped_C2C_from_ST2_ACT_to_ST2_CPS_2014_04_01.fig

/ligo/svncommon/SeiSVN/seismic/HEPI/H1/ETMY/Data/Figures/Transfer_Functions/Simulations/Undamped
H1_HPI_ETMY_TF_C2C_Symmetrized_from_ACT_to_IPS_2014_03_21.pdf

Grab the lowest-frequency value where these are flat.

(P_PS)
       Z [nm/ct]   RZ [nrad/ct]
HPI     14.70        6.80   
ST1      2.90       13.45
ST2      1.09        2.70

IPS_ST0 = z_ST0 - z_GND
   => z_ST0 = IPS_ST0 + z_GND

CPS_ST1 = z_ST1 - z_ST0
   => z_ST1 = CPS_ST1 + z_ST0
   => z_ST1 = CPS_ST1 + (IPS_ST0 + z_GND)
 
CPS_ST2 = z_ST2 - z_ST1
   => z_ST2 = CPS_ST2 + z_ST1
   => z_ST2 = CPS_ST2 + (CPS_ST1 + (IPS_ST0 + z_GND))
       (Assume z_GND == 0 over long-term trends.)
   ==>> z_ST2 = CPS_ST2 + CPS_ST1 + IPS_ST0
        z_ST2 = P_CPS_ST2_Z * ST2_ISO_OUT_Z
                + P_CPS_ST1_Z * ST1_ISO_OUT_Z
                + P_IPS_HPI_Z * HPI_ISO_OUT_Z
   
(Same logic flow yields the same for RZ)
   ==>> rz_ST2 = CPS_ST2 + CPS_ST1 + IPS_ST0
        rz_ST2 = P_CPS_ST2_RZ * ST2_ISO_OUT_RZ
                + P_CPS_ST1_RZ * ST1_ISO_OUT_RZ
                + P_IPS_HPI_RZ * HPI_ISO_OUT_RZ
                
                
BUT ALSO, can use this same trick to calibrate ASC outputs from the M0 LOCK banks 
w/o knowledge of the ASC loop gains or crossovers (again assuming only that they're DC coupled over hours-to-days time-scale).

OPTIC PIT/YAW = P_(M0 CTRL_to_OPTIC) * M0_LOCK_OUT

P_(L3/M0)

Found from 
/ligo/svncommon/SusSVN/sus/trunk/QUAD/H1/ETMY/SAGM0/Data/
    2021-11-02_2156_H1SUSETMY_M0_Mono_WhiteNoise_P_0p02to50Hz.xml
    2021-11-02_2156_H1SUSETMY_M0_Mono_WhiteNoise_Y_0p02to50Hz.xml
    
Plotting the not-usually-plotted, but present in the template TF, M0 EXC to L3 OPLEV
Took the lowest frequency point, (checked for good coherence, of course)

Both are coincidentally 0.03 [urad/ct]