JohnRob:
The Magnetic encoder is a fine solution. They are a little costly (compared to a optical/contact "O/C" rotary encoder).I've not used the magnetic encoders but from the datasheet they are a serial interface, and may be easier to program. They also seem to have a very high resolution compared to the O/C encoder. And they provide an absolute indication (i.e a pointer on the knob in a known position will always provide the same reading). The O/C is relative like you find on an modern automotive radio volume control.
From what I read they don't have any detents (as expected with 10 bit resolution) so you will have to provide some friction to stop vibration from changing the angle.
I have used the O/C type encoder for applications where didn't want to sense "noise" I found a 16 position (4 bits) O/C encoder perfect for my operation. I don't recall the mfg or model but it had detents at each position so I didn't have to worry about changing from vibration.
Keep in mind you are not limited to 16 positions as the encoder can be turned more than one revolution.
Also you can define where ever it is on power up as "0" and not worry about starting up with the control at max.
From what you described, i think that the magnetic encoder will actually be the best fit for what i am looking for, as i will not be able to make a full rotation so high resolution is a must. Also along with being constrained to a small angle of movement, roughly 130* based on how i build my handle, knowing the absolute position of the handle will be a better fit for this project. As i will want zero to all ways be at the minimum of the throttle handle position. Also i think you are right with the respect of being a bit easier to code, it should also allow me to read the position of the handle when i choose to as opposed to having to use an interrupt to catch movement when it occurs.
Thank you for you input.