I carried some further modifications and tests to the AI Board. Details and observations here:

• I switched out the resistors for all the remaining 7 channels, using the same substitutions as detailed here. 
• I then verified that the modified transfer function for all 8 channels using the SR785. I did not collect data for all the channels as netgpib was taking ages, but I did use the cursor on the screen to verify the position of the first notch at ~64 kHz. I noticed that all the channels did not have the lowest point of the notch at the same frequency. Rather, (at least on the screen), this varied between 63kHz and 67kHz. I would put this down to component tolerance. Assuming 5% tolerance shifts the theoretical notch frequency from 66268 Hz to 63112 Hz. 
• After verifying the transfer functions, I went to 1Y4 and plugged the AI board into the eurocrate. I then connected the input of the AI board to the DAC output using my custom ribbon cable. Next, I used the excitation points set up earlier to send a 1 kHz, 32000 counts amplitude sine wave through the channels one at a time. I monitored the output using an oscilloscope and the LEMO monitor channels on the front panel of the board.
• I found that the single-ended output of the AI board swings between -10 V and 10 V (w.r.t ground, oscilloscope trace attached). This is good because this is the range of input voltage to the PZT driver boards required to realize the full actuation range of the PZTs.
• I also verified that the connections on the custom ribbon cable are correct (channel map was right) and that there were no accidental shorts (I checked other channels' output monitor while driving one channel). 

I think the board is okay to be used now.