There is a lot of conflicting information on what the CO adjustment does. My current understanding is that it adjusts the base fuel injector "on" time by a set amount across the board (not a percentage). If that is true, then the effect would be most pronounced at low throttle openings, diminishing with increased throttle (and increased injector "on" time). In closed loop running, the O2 value drives the ECU to correct the base injector "on" time to give a stoichiometric mixture via short term fuel trim calculations. So in closed loop running, the CO setting should make no difference. Add to that the strong likelihood that the S10 includes both short term and long term fuel trim values, just like most automobile systems do. The long term trim values would be adjusted when the short term trim values remain for several seconds. Long term trim values would likely be applied in open loop running as well (they are an indication of something in the fuel delivery system being non-optimal). It would appear that the CO setting is meant to compensate for things like fuel pressure that could vary from unit to unit, to keep the fuel trim levels in an optimal range. With all of the fuel mixture manipulation that the ECU does, it becomes difficult to predict the outcome of extreme changes to the CO settings. It's not the same as just putting in a larger jet in a carbureted bike.
I could be wrong on what Yamaha's fuel injection algorithms do, but it seems reasonable to assume they do what other systems do.