As you know, I am in Ontario. And even here, I know I would be hard pressed to find a truly knowledgeable Tube Tester Technician. Of course, I myself do not know, if the last Tech who worked on your Tester, or as to what he knew or what may have messed up.
Most of what I am saying here, I have learned from restoring my Jackson 561 Tester. Also do you have a schematic for this Stark 12-22. While your Stark is more complicated than my 561, it should not be too hard to do it yourself. Just like a radio, replace all the caps, test all resistors, clean all the switches and pushbuttons, Potentioameters/Rheostats with Dioxit. Clean all tube sockets real good, and if you have another Tube Tester you can really trust, check the tubes used in your Stark Tester, including cleaning the pins. PS: When checking resistors, remove one end of the resistor to ensure nothing else in the circuit will influence the reading.
As line voltages change during the day as loads placed on the grid can change, I would use my tester on a reliable Variac to ensure that I am getting reliable Filament Voltages on the Filament Switch. This is important when seeking optimal working conditions for the tubes being tested. It will effect other necessary voltages also.
Also let the tester warm up of ten minutes or more. Also too, make certain the tubes in you unit are up to snuff. This also applies the the Meter itself. A faulty meter guarantees you have a Tube Tester which is not accurate.
However if the meter is just a little off, you will soon get to know its limitations, and will be able to compensate for its shortcomings.
Do you have a copy of the Operating Manual for this Tester. Also do you have a proper Schematic. These machines are not Rocket Science, in fact maybe easier than most radios you work on, so there is no reason why you should not be able to get this tester into reasonable working condition, barring a blown meter, of course.