jclough :
The wheel balancer didn't take off because it didn't work!
But it was patented - just shows you how much that is worth sometimes - and as I recall, the patent did not mention critical speed. It definitely had balls in it, not mercury, but the name "balancemaster" seems familiar. Enough mercury to balance four car wheels would also cost a fortune and give the EPA guys heart failure. Maybe the mercury has damaged the inventor's brain (just kidding)! Devices like these also require some precision in regard to the concentric positioning of the tube and the center of rotation. There have been a number of active autobalancers developed over the years, (rather than passive ones like you are considering) but they are all rather complex. I believe I read that there is an ongoing investigation to incorporate them in jet engines, but my memory could be at fault.
I personally don't think your spiral vane idea would work, although I may not understand it correctly and could be wrong. Is your fan horizontal or vertical? In the horizontal case, it's not clear what would keep the balls at the center of the vanes when stationary. In the vertical case, what would be the function of the vanes, and what is the function of the spiral? The balls would get flung right to the outside immediately, with or without vanes. The balls have to take up certain angular positions all at the same radius, and hopefully remain there. I repeat my suggestion that I think you need some means of mechanically locking them in the most recent optimal position until you are close to the operating speed (which I think is the way the Cincinnati grinder worked). I don't see why this should be so difficult. Whether it's worth it is another matter, as others have said.