Another source for ideas are papers in large HCI conferences, e.g. ACM CHI. Researchers have prototyped and evaluated lot of input devices in many variations.

8-10 dimensions is a lot, to the point that I'm not sure an input device that allows to control all of them simultaneously is viable.

There are so-called "space mice" which are often used by people working on 3D objects and have many degrees of freedom.

Various tracking devices (Kinect, Leap Motion, Wiimote) can theoretically be used to get a lot of inputs, but when moving in "free space" it's very hard to just adjust a single factor independently (if that's required).

Also, is very hard to work efficiently with two hands on independent parameters: generally it's more efficient to have one hand generating value input and another switching parameters. Again, there are tons of papers evaluating this kind of thing.