That's how it works over time too... it's the trends you watch for... even if it is still inside the normal range, if you see it creeping up/down over the years, you should watch it. Even if you are naturally high or low, if you trend into the normal range, that isn't good either because you aren't naturally at those levels.
Problem is that the creep may happen in small enough increments that you might not notice it without graphing it out. Mostly, I can't get a good grasp of it seeing it as numbers, but on a graph i can just look at the dots and see if they are trending up/down. Some of it is just getting older, and nothing to worry about, other might not be. If you go to the doctor on a regular basis, he is looking at it too (unless you shop doctors and they might not have years of history on you).
Most of the things in a basic metabolic should stay fairly constant, same with a hemogram (but even that isn't "constant" in that stress affects them too) Not enough to cause it to be worrying but they can jump around and if you just look at it yourself and see oh, my wbc just doubled from 5 to 12, I have a problem... maybe, maybe not, could be just allergies or something or you pulled a few late nights prior to the test. But they are still in the normal range too. Liver/Lipid tests jump around when you eat and other things, these are useful but again, the numbers by themselves while good for treatment isn't good for diagnosing on their own because so many things can affect them.
The idea of monitoring your own results is fine, it's just it causes unneeded stress too when you start to self diagnose. If going this route, you could just look at trends, and if you notice it, then at the next annual appointment just bring it up with doctor and ask him about it. Unless you were told to watch for something, it most likely is something that can wait until then, so you don't need to rush and call an emergency appointment with him.
Is there a reason you don't want to use a real lab for this?
I'm fairly certain the labs the the results are coming from are real... it's more of a problem of not having a real expert interpreting it...