I've been experimenting with different layouts for how to retract the optics.

My original idea had the mirror assembly swinging in line with the camera/nozzle and away from the nozzle.  That required a swing of a full 120 degrees with a best case no gearing time of 0.34 seconds to deploy.  I was not happy with that.

That was replaced by swinging it 90 degrees to the side, which then resulted in a best case time of 0.25 seconds.  I was still not happy..

The last iteration has the mirrors back to swinging inline with the nozzle/camera, but now swinging them towards the nozzle.  That only requires a swing of 60 degrees with a best case no gearing time of 0.17 seconds.  Only issue with this design right now is that the pivot point required is inside where my lens was going.  The lens is the gray cylinder here.

I had one more idea come to mind with both mirrors on an uneven length linkage, and then they could both move different directions out of the way.  Problem is then I have to generate two extremely reliable meshing linkages.  The mirrors have to stop at exactly the same spot every time or the part centering will not be reliable.  I'm not pursuing that one for now.

I've also switched to planning on printing this on it's side so some things changed, like the ability to hollow out the middle without generating bridges and using hexagonal ends to avoid overhangs.  This makes fixing the 'middle of the lens' pivot point a bit more challenging though too.