Thanks for your interesting comments. To pick up on a few points:

Balance
Both reverett123 and Sasha mention this. I see the importance of measuring this too, but it's one of those things I find too difficult to get reliable readings for in a reasonable time: either I'm too good (left side, eyes open, indefinite) or too bad (right side, eyes close, almost zero) for a simple test like stand on one leg for 60 seconds to mean much. As Sasha mentions, the pressure pad approach might be useful - perhaps pressure variation could be measured.

Using a computer
Again both reverett123 and Sasha mention this. It would be especially interesting if tests were standardized and web enabled. Ronhutton mentions reaction time testing. This would be easy to computerize.

Existing tests
KC Tower mentions the Purdue Peg test and other people mention other tests. To stop us reinventing the wheel and make communication and aggregation of results easier, it would be useful if PwPs had a list of standardized tests.

Other factors
Paula W is right to mention the importance of psychological factors. This must include the placebo effect. The only way to reduce this that I can think of is to have a larger sample size.

Hospital based measurement
Ronhutton mentions having tests every 6 weeks. Is this part of a research project Ron? Do they aggregate the results into a single index or, perhaps, a few orthogonal indices?

Usefulness
reverett123 recounts on another thread (93463) [I had a link here, but I've had to delete it because my post count is less than 10] the effect of potassium on his PD symptoms and the limited research in the area. If only we had the data!