I think I might agree with your suggestion but for one small detail. Is whatever is being sensed a one for one correspondence with the magnitude scales? If they aren't, I wouldn't think it should be done that way. But I may be too generous, but I also may be to restrictive.