I have spent my career as Igor working for Dr. Frankenstein who was doing science. It has been a frustrating experience. I am an engineer and think like an engineer. Cost is a major part of my thinking. I don't try to do the best possible, I try to create the most cost effective solution to meet my objective.
Defining the objective has always been a problem. When I go to Dr. Frankenstein and ask for specifications for the lift to move the body to the top of the tower, I always get a vague answer. "We may want to try to lift an Elephant, or if my DNA experiments are successful then we will need to lift a Brontosaurus" he says. I say "But master, if I need to raise a Brontosaurus to the top of the tower I will need more money for the lift" In fact, a Brontosaurus lift is not very practical for lifting a standard model B big brute body. I have found over the years that if I listen too closely to Dr. Frankenstein's specifications, I build worse equipment for the experiment than if I designed the best I could within the practical constraints of technology and budget. I tended to design at the knee of the cost curve, even if that did not provide the best possible measurement.
So I have always seethed in frustration during the design process. I have been at this for 40 years. At Princeton and Fermilab I was given a lot of freedom. The way I learned to cope was to take time to understand what the experimenters were trying to accomplish. To their defence, they do not know what is important in an experiment. That is the definition of research. If you knew the answer you would know what was important. Not knowing the answer, one must guess. Scientists don't like to guess, so it is hard to get answers. When you do get an answer, it is always a Brontosaurus.
At Princeton I ran the equipment pool on the side. So I could observe which NIM modules were the most coveted, and which got used to do an experiment if they were all that was available. I thus learned to differentiate what was needed from what was nice to have. I would then sit down and design a good economical instrument and make a few available for test use by the experimental community. Sometimes nothing happened, sometimes they got designed into an experiment and then I had a production problem. If I did my job correctly, the latter happened more often than not.
When I designed the Mark IV, I listened a lot to the people that had been attracted to the Mark III project. I particularly read the various versions of Bohden Paczynski's papers on variable star searches. I then designed the Mark IV as a search instrument. It is not a "measure and characterize a star" device.
The Mark IV is thus designed to be a survey device. It is designed to cover the whole sky as often as possible in many epochs. It is designed to mark stars as "probably variable". One would like it to do everything. But that is the engineering compromise.
This is the result of letting Igor plan the scientific program. If you wanted it to be different, then you should have talked more persuasively to me a few years back when I was doing the design. I sure did it in public, so anyone who cared had an input.
So John, the answer is "mark as many stars as you can as probably variable for which there is evidence in the survey".
This may not be the most desirable scientific result, but it is what the Mark IVs were designed to do.
Next: How I plan to Operate My Mark IVs.