My first suggestion it that your beakers are upside down in the first photo - stop working on the cieling! All the solution will fall out! ;^)
Don't worry about the interface box - a modern VOM should have high enough impedence that it will work fine. Keep things simple. And don't put a pot in there - it's not needed.
Stir the test solution so that fresh solution is being brought to the electrode continuously.
Also, try replating the electrode - I think it took me a few times to get mine seasoned so that it worked better.
And most important - remember that reference electrodes are half the equation. If your reference electrode doesn't match everyone elses reference electrode, then you will have a fixed bias, positive or negative, from everyone else's results.
Make a series of solutions with a log series of concs, like say (and just pulling numbers from my head) 0.03M, 0.1M, 0.3M, 1.0M, 3M and then measure those. Plot them on some Gran's plot (semilog) paper and see if they are linear. If so, then you have a functioning elelctrode. If there is a bias, then it could be the vAg electrode or the reference electrode. If there is no linearity on the Gran's plot paper, then something is not working as it should.