Thanks Kevin.
I experimented with the opto server tonight (this time with the OP27 in place) and here are my observations:
Without the servo jumpers in, I first adjusted the offset of both boards to be about +18 VDC at cold start, measured between the + output and ground (this applies to all subsequent measurement).
I then inserted the jumpers and powered on the Carbon with opto servo now engaged. The offset on one board measured -2.5 VDC right from the start, it dropped to about -1 VDC after 5 - 6 minutes, from there it hovers around -0.9 VDC, drifting about 0.5 VDC or so up and down. I measured again at 20 minutes and then at one hour marks with same results.
At the one hour mark, I adjusted the balance from around 2 VDC to within 0.5 VDC (absolute value). This brought the offset to hover around -0.1 VDC.
The other board exhibited similar behaviors but started with and ended at lower offset voltage. (Without servo, this board has always had lower offset and balance compares to the other one.)
It appears to me that the opto servo is working and it's effective from the very start although it does take about 5 minutes or longer to reach the lowest possible offset voltage.
As I have mentioned in the past, the Carbon, without the servo, is very stable to start with. It seems to me the main advantage of the opto servo is to keep the offset at a low voltage from the start and settle it down faster. Without the servo, my Carbon typically takes between 15 - 20 minutes for the offset to settle.