You've invented a non-equilibrium in which CO2 biosphere-ocean sequesters are saturated to the point the system cannot handle a 0.09% increase in total CO2 content.
Why do you insist on expressing the increase of CO2 as a fraction of the total amount of CO2 on the planet, including what is trapped in the oceans, rather than of the volume of CO2 that is currently in the atmosphere? Do you believe that it doesn't matter into what part of the cycle we put the CO2, the effect will be the same? Are you saying that if, instead of putting the CO2 directly into the air, we could dissolve it in the ocean and it will be the same thing? Do you not see the flaw in this logic?
Let's take my vessel (the atmosphere) example from before and expand upon it. Replace the fixed-rate bilge pump with a hole in the bottom to represent the rate CO2 is re-absorbed by the ocean. As the vessel builds more head, more water should flow out of the hole, right? Is this a reasonable way to imagine how the earth might increase CO2 absorption as atmospheric levels rise? Might not be an off switch, but it is a dimmer. Now, in addition to our "atmosphere" vessel, we have a much larger "ocean" vessel. It is similar to the atmosphere vessel, except the ocean vessel drains into the atmosphere vessel. Whatever drains from our atmosphere vessel is then put back into the "ocean" to complete the cycle. At the start, both vessels drain about the same rate of flow that we will say is 1 liter per minute, and at this time the atmosphere vessel has about 1 liter of water in it. One liter comes in, one liter goes out. The level in the atmosphere stays the same.
Now let's imagine that as we put each liter that drains from the atmosphere back into the "ocean", we add 30ml to it at this point in the cycle. As you said, since 93% of the total water (CO2) in our system is in the "ocean", the ocean level will be barely impacted, and same goes for the drain flow as a result of this minutely increased pressure. Barely more than 1 liter per minute will drain into the "atmosphere", and the level in the atmosphere vessel will rise very slowly, if at all.
Now instead of adding those 30ml to each liter as we return it to the "ocean", let's put those 30ml directly into the atmosphere vessel with each liter that naturally drains into it. Do you not see how the atmosphere level will rise much more quickly if it is added at this point in the cycle? It's the difference between, as you keep pointing out, tipping the scales by 3% vs. by 0.09%.