I think it's plated correctly, but this was a first attempt. I think I will clean it and try again to see what happens. The meter does have a high impedance input. I borrowed it from the lab at work and it was just calibrated so it ought to be correct. Besides, the opamps in the buffer box use j-fet inputs IIRC and have in input impedance on the order of 10^8 ohms! I get the same reading with and without the buffer.
How much would concentration of any of the solutions have to be off to make that sort of error in the reading? I am guessing pretty far off for the sort of error I am seeing.
Probably would be a good idea to check out the homemade reference electrode as well, once I figure out how to do that. Open to suggestions!