This is how the calibration of the DIY meter looks, for the first version, with the sensor head inside an acrylic jar. This should be repeatable if anyone else does one this way.
The next test is the one it will probably fail - submerging it in water to see if it leaks.
EDIT: First a 5 minute dunk, then a 15 minute dunk. No leakage that I can see! Next will be to compare the readings in water to the PAR meter, to see if the calibration changes.
EDIT: Tested in aquarium, 2 feet high (2 feet head of water pressure on sensor). The data point falls on the same line as when calibrated in the air. This definitely works. However, the sensor is still too buoyant, floating up if your fingers slip at all. And, it needs a "wand" to be able to use it effectively without your hand getting in the way.
I think the sensor-from-scratch that I'm making will be a big improvement.