I wanted to add a dosing reservoir to the original design. So I replaced the Tee with a 1 cup jar. The dosing output is at the bottom of the jar, and the air input is in the lid of the jar and the dosing input is also in the lid of the jar.
I originally tried this with the jar below the dosing container, but I kept getting trapped air and a continuous siphon of dose into my test vessel. I don't know why this is, since the test vessel was well above the dosing container.
The latest design has the dosing reservoir inside the dosing container (weighted down with rocks) and the same connections, except that the Check Valve into the dosing reservoir sits directly in the dosing container.
This endeded up being very similar to the dosing tubing in the dosing container design that I saw elsewhere. Except that my Checkvalve prevents the air from blowing out the refill hole when the air pump is on. I may remove the checkvalve from the dosing reservoir input hole to see how much air is actually lost by allowing two way flow through the input hole. I suspect that most of the water that refills the dosing reservoir is sneaking in through the lid on the jar and not through the checkvalve (testing required). I did have the lid siliconed shut when I was getting the trapped air, but I had to remove it to add my buoyancy counterweight (rocks).
The good news is that it seems to be working as planned now. I just need to make sure that no fluid is leaking after the timed event. I also want to experiment with the refill design as mentioned above. I used RC Aircraft fuel barbs for my inputs/outputs and 3M 5200 sealant to seal them in place.