Nevermind the meaningless blather

So, after trying so hard to figure out the distribution functions, I realized this:

My program could accomplish one of two goals:
1. Tell us what we are going to see
2. Tell us how efficient we will be at detecting what is actually there

If I wanted to predict what the survey will return, I would need the distribution functions like I have been trying, but if I change the goal to merely assess the efficiency of the survey, I should just use a uniform distribution function. That way I can tell how the distribution function changes after experiencing the noise and detection. Maybe later I may run the simulations to predict what we will actually see, but for now I will continue writing my code the simpler way.

Later: Meh, that itself was just substance-less blather... If I am just going for efficiency, then the code will probably just tell me to look at a few things really well. Number of planets is the one way to strike a trade off between efficiency and actually managing to detect a wide number of planets. How would I factor size of sample in to Goal 2?

Further complicating the Goal 1 way to go about things, the distributions that I got yesterday were post-biasing. I need a pre-biasing distribution so I can then bias it and see what happens. But I don't know the parameters of each individual survey to get a "bias surface" to un-bias the surveys.
I don't even know if my weird concept of a "bias surface" exists. It would give the percent observed out of percent present. So simple conversion from the post-biasing distribution function would give the pre-bias distribution surface. Probably requires many more data points than those available.

OH OH!!! Just found paper on survey optimization. Hopefully that will solve my problems.

Survey Update
M44: 1 new red subgiant at 6.68 mag and 2.2 solar masses (4 others eliminated)
NGC6404: 4 9th magnitude possibilities