Hi Lee,
Stable CO2 can be thought of as a concentration level that is sufficient for the available light that is more or less constant throughout the photoperiod.
Carbon fixation is a complicated procedure which involves many chemical reactions but the most important reaction is the first one. This involves the capture and transport of the CO2 by an enzyme called "ribulose-1,5-bisphosphate carboxylase/oxygenase" shortened to "RuBisCO".
The production of the enzyme RuBisCO is tightly coupled to, and controlled by the concentration of ambient CO2. For "X" amount of CO2 present the plant produces "Y" amount of RuBisCO. This maximizes the efficiency of carbon fixation. If the quantity of available CO2 suddenly drops the plant then has to reconfigure it's chemistry to raise the amount of RuBisCO. Similarly, if the CO2 rises, a RuBisCO production decrease is accomplished. This would be OK if RuBisCO production and reconfiguration in higher plants were quick, but as it turns out, this procedure is very slow - a week or two may be required. Evidently, some plants adapt more easily than others.
You can see therefore that a fluctuating CO2 level causes a rise and fall of RuBisCO production, and during this time the photosynthesis machine struggles to produce the sugars and proteins required for growth while the plant is busy attempting to adjust the machinery to the CO2 levels.
This is only relevant of course during the photoperiod. None of this matters during the dark. If you turn the lights on and the CO2 levels are low, this yo-yo begins. The morning is more important because this is when the plant transitions from 100% respiration to food production. Near the end of the day the plants start to "close up shop" so the levels are not so important then. You can often see this behavior on plants like L. aromatica or D. diandra as they start to fold up their leaves near the end of the photoperiod.
The trick is to have high stable CO2 levels by the time the lights come on but it's OK if the levels slowly trail off in the afternoon. Remember that the plants are removing CO2 from the water so enough CO2 has to be added to compensate. The easy way out is to have 24/7 CO2. This works fine and doesn't require any acrobatics with a solenoid. The disadvantage is that you can't get as high a level of CO2 during the photoperiod without killing the fish and the CO2 consumption is wasted. The fish can also have a problem in the morning when oxygen levels are depleted and CO2 levels are at their highest.
Algae have no trouble adjusting to fluctuating CO2 levels and it may even be that fluctuating levels help to trigger some algae types such as BBA. One of the first indications of lower than expected CO2 is the appearance of green hair algae. I see this when I'm not paying attention and the CO2 cylinder is gets low causing an injection rate falloff.
I either have to adjust the needle valve or swap out the bottle and/or temporarily lower the light. Unfortunately it then takes weeks to stop the hair algae production. I don't panic though because I realize that the RuBisCO production stabilization takes time. I just get the toothbrush out and twirl away once or twice a day. It's when you misread the signs and take inappropriate action that the trouble begins. Some people panic and decide to stop dosing, which is of course completely absurd.
Hope this helps clarify.
Cheers,