Interesting perspective gloe, you have me thinking that a range chart may also further distort the indicator than time based bars. That being said one can see from the chart of the 15th how spikes in volume during both the London and New York sessions cause the CO to distort giving inappropriate indications. I imagine the same would ocurr on a time based chart. If one used a volume based chart you would effectively be turning the CO into a PA based indicator only and thereby negate the value of it's volume input. I don't see a solution other than running a volume indicator in addition to the CO, do you?
BTW it can be seen from the chart below how the CO amplitude is directly proportional to the volume. If one believes that volume precedes price then the CO will work but I have little faith in volume as a leading indicator.
Market Data 
Markets 
__3_15_2012.jpg)
AI
