lnewqban,

The Reynolds number for the transition from laminar to turbulent that you posted (2100-4000) is valid for pipes and enclosed flow. This is not the situation we see in an airfoil and the transition occurs somewhere in the vicinity of a reynolds number of 500,000. This is of course assuming a perfectly smooth surface, which is not really valid because no such thing exists. So in reality, the transition will be before 500,000, but the point is it is a lot higher over a plate or an airfoil than flowing through a tube.

That being said, the water test is actually very much so valid, if you do it correctly. Now, I do not believe that the OP did the test correctly, so his results are most likely wrong. But if he takes the time to set up the experiment, then you can get very close to how it would perform in air. Using water you will be able to compare different airfoils very easily as well. The numbers might not be exact, but all the relationships will be present. And even though this may not yeild exact and perfect numbers, it would be more than adequate for the world of model airplanes.

Now if you really want to have some fun, get some glycerin and do the tests in that. The viscosity is huge and you can really slow the flow down and see what is going on. One cool trick is to take a syringe filled with some colored glycerin and inject it into the flow of glycerin and you can see the streamlines around the airfoil.

These small models can do things that real airplanes can not, so rigorous testing of an airfoil that you will simply put on a model is pointless to me. Unless you have a friend with a big windtunnel, then give him a call. But I would not pay to use a windtunnel or build a windtunnel or anything like that. Seems like a waste of time to me, that is less time you have for flying!

Point is, a test with water is valid if used correctly, and a test using a windtunnel would be ideal, but not absolutely necessary for what you are looking for.