Word definition in the following discussion 

the best DARM -> DARM with MICH FF gain =0.1, PRCL FF gain = 0.2 on 9th March,

worse DARM ->  DARM with MICH FF gain =1, PRCL FF gain = 1 on 9th March,

Discussion:

~19Hz peak, ~23 Hz peak, ~50Hz peak, ~58Hz peak, ~74Hz peak ... and structure from 100 Hz to 200Hz in PRCL is similar to the "best DARM" . 

While the structure for almost frequency range except for several peaks in MICH is quite different with the "best  DARM".

However the structure from 100Hz to 300 Hz in MICH is quite similar with "worse DARM" before we  reduced the MICH control loop gain.The gain reduction of MICH FF is 1/0.15. 

Question:

Did these MICH (green) and PRCL (sky blue) includ the MICH FF loop gain reduction ?  If so, this can explains very well why MICH FF loop gain contributed to DARM noise reduction especially from 100Hz to 300 Hz and  why PRCL FF itself did not contribute the noise reduction in DARM.