Great strategy, Levers101!
I'm not sure I'm with you on the pH rising as the yeast no longer produces gas. Wouldn't it just stay much the same as long as the carbon dioxide "blanket" remains in the vessel?
However, I agree that measuring the gas phase is not the way to go.
Now here's the basis for an idea for those of you still interested in density measurements:
- A hydrometer reading is just a level (height) reading which we convert to SG by means of a scale.
- There are electronic (optical) instruments for reading levels in tanks etc.
Make the instrument in the shape of a hollow cylinder with thick walls (open in both ends). It needs to be weighted so that it will "stand up" in the wort, like a regular hydrometer. Now put a string of, say, regularly spaced light sensors along the bore. (Photo diodes or some such might do the trick.) A similar string of LEDs could be placed on the opposite side of the bore, or you could rely on ambient light.
Now, above wort level, the light should have no trouble reaching the sensors. Below wort level, the sensors should detect much less light (this probably needs calibrating for a pilsner.) The voltage from the string of sensors would translate to a level reading, which can be calibrated for two points on the SG scale -- say pure water at 1.000 and a sugar solution of 1.100.
I don't know if this makes any sort of sense at all, but I've even drawn a diagram. It shows the top, cylindrical part of the instrument. The rest of the "payload" (temperature measurements, electronics) could be kept as ballast below the surface.
(You wouldn't necessarily need light detectors. Just a string of something that reacts differently to water than to air. Preferably something with an electrical output. Resistivity probes might do well.)