I don't think that method of measuring thrust would yield very accurate numbers at all no. Way too many variables involved there. Only way I can see you really getting an accurate reading is to hang the plane hanging nose down from the scale somehow, you would also want the scale to be exactly in line with the thrust line vertically. Then of course run up and quickly cut the motor when done measuring thrust there. After motor's cut then subtract the full weight of the airframe (Remember you don't want to run out more fuel before measuring the actual weight then). This whole method is obviously going to be very dangerous regardless how you do it and yes very unpractical. Unless you can come up with a contraption to do this. So ya in the end to sum it all up your method of doing it already will give you a ballpark figure to work with but not the most accurate.

For your method I would suggest a really smooth surface to do it on to begin with so the wheels can really rotate freely, and also still try to get the scale to be exactly inline with the thrust line! Another thing to avoid while using this method is any other forces on the scale like picking the plane up slightly etc. Very difficult.

To answer your other question, the weight of the model will have no effect in the perfect enviroment. But yes like I said above if it's for instance a tail tragger and you're picking the tail up by the scale somehow or whatever you could get some different end readings on the scale.

Another point you should bare in mind as that static thrust measurements is not going to reflect what is actually happening upstairs! You can maybe measure your thrust readings when hovering sure (Yes it's the 3D forum I know so it is applicable there). But we have to remember that there are other factors involved like wind even when hovering which will make the prop tend to become a little more efficient (This theory I base on my heli flying experience and it's just a theory). Also you have other factors like prop cavitation etc which will change as the model gains forward speed depending obviously on the prop. Example: it's impossible to to do thrust and amp draw readings on my little brushless pylon racer because it is swinging a 5X5 prop and it just cavitates on the ground, upstairs when the model's moving it draws MUCH more current as the prop gets enough air supply to bite into.

I'm by no means an engineer. And yes I'm sure someone will chime in sometime that has a good method of measuring thrust statically. Your method of doing it is really what I would do also to just give me an idea of what the motor is capable of when the model is hovering perfectly still.

Cheers
Spacey