It doesn't, assuming all of the nosepiece fits into the 2" focuser receptacle.
I've measured the effective optical paths of a few diagonals myself. The 2" model with the shortest optical path I have come across so far is the Baader 2" BBHS Zeiss-spec Prism #2456117. I measured it on two different occasions and got 96 mm to the top of the 2" Clicklock clamp. My estimates of optical path are probably accurate to +/- 1 mm.
Hmmm - I think we're discussing separate issues. This topic (as I understand it anyway) is about converting a regular refractor into an instrument that has properties similar to a NP101. Specifically, around f/5 - f/5.5 and capable of using 2" eyepieces so that the true visual field is maximized.
I tried this with my APM 115 (natively f/7.0) and TOA150 (natively f/7.3) some time ago. I used a TOA reducer. I found that - even though both instruments were designed for imaging - the primary issue was achieving focus because the light cone from the reducer converged too rapidly. Solution - bring the eyepiece closer to the reducer.
1. The only variable in my system was the distance between the reducer and body of the diagonal. It may be different for other reducers, but my TOA130R has its last optical element 6 mm from the rear rim of its flange. This means the system maximizes the available back- focus for imaging.
If you were to install a nose piece on the diagonal, you would need to install a 2" receptacle on the reducer which pushes the diagonal about 30 mm from the reducer. If the diagonal is screwed directly to the reducer, you can do this with an adapter that only eats up ~3 mm.
2. I realize that the optical length of a 2" prism is ~15 mm shorter than a mirror. The problem in this specific case (f/5 - f/5.5) is that a prism creates a lot of spherochromatism. This is why I did not suggest a prism (even though I find my T2 BBHS prism produces my highest contrast shallow space images.)