I'm not sure how you're doing it already exactly, but rather than assign a single "pan" value I'd have two audio tracks: make both identical. One 100% for left ear and one 100% for right. Modify the volume rather than the pan. That way, rather than simply having it move from ear to ear, it can have a mix of balances, presumably with both at 50% volume so at its loudest it would only be at a standard volume. I know I keep plugging Mixcraft but I don't have much experience with other audio programs and it's pretty simple to do--you're effectively just adding "anchor points," so to speak, onto the track itself and you can assign volume values to the hundredths.
Going further than that, it's hard to say what the easiest thing to do to convey three-dimensional movement would be, since panning only does left and right. If you want to get even more complicated you could have four tracks:
1. LEFT, Equalizer - Cut Bass
2. LEFT, Equalizer - Cut Mids & Highs
3. RIGHT, Equalizer - Cut Bass
4. RIGHT, Equalizer - Cut Mids & Highs
With the basses effectively being "backs" and mid & high tracks being "fronts," you could do four-point movement. That would probably take an ungodly amount of time to do properly though.