Here's a copy of a post I just put on over at the SL55 Forum.
Originally I was looknig for rebuild advice so I could get a cheap 2300 and refurbish it, but now I'm really wondering about the gains.

..........
What I've learned so far is that both the 1600 and the 2300 are very similar except for...
1) The front snout of the 2300 has an additional gear set in it to drive the main rotor to; 6200 x 155/91 x 1.35 = 14,288rpm + the meshing gear to drive the 3 lobe rotor to 5/3 x 14,288 = 23,813rpm
2) The front snout of the 1600 on the SLK32/C32 only has the mesh gear - but the main rotor is driectly driven by the pulley to; 6200 x 155/74 = 12,986rpm + the meshing gear which drives the 3 lobe rotor to 5/3 x 12,986 = 21,644rpm.

So you can see the 2 have very similar specs in relation to the bearing speeds - so I would safely assume they are of very similar construction. I was originally under the impression the 2300 was spec'd differently, but I'm not thinking that any more.

When you work out 2300 cc/rev versus 1600cc/rev
I'm currently assuming the /rev is per rev of the pulley shaft.
If you assume 1 rev of the main 5 lobe rotor (female/cavity) does this 2300/1600cc
1) for the 2300 you only have to turn the pulley 1/1.35 times to make one rev of the female rotor
2) for the 1600 you have to turn the pulley 1 time to make one rev of the female rotor.
So the 2300 has an internal 1.35x advantage.
If it didnt have this internal gearing, it would only do 2300/1.36 = 1700cc/rev against the 1600 which does 1600cc/rev.

The other key difference I've established is that looking at the 2 units in the engine bay, the 2300 has the teflon coated male/3 lobe rotor on the other side (drivers LHS). This is because it is driven through the additional internal gearing set in the snout, whereas the 1600 has it's 3 lobe male rotor on the drivers RHS because the pulley direct drives the 5 lobe rotor.

Now for constructability - if I was manufacturing these for profit - to minimise costs - I would use the same rotor profile but just cut one longer/shorter to fit.
This leads me to belive there is maybe only a 10% difference between the lengths of the 2 rotors - the 2300 being 10% longer.

Externally - I measured my rotor case at 220mm and from comparable scaled photos I've measured the 2300 at 278mm. But I havent taken into account bearing beds and the like - hence why I am chasing the actual measurements to confirm the hypotheses above.

It was all very interesting 'cause I started out fantasising about the effect of replacing the 1600 on my SLK32 with a 2300. I was thinking efficiency improvements and air flow. But if they are very similar rotor designs/profiles, the gains would be minor. The main difference between the isoentropic charts would be that the 2300 is simply a 1600 stretched out by a ratio fo 1.35 + 10%.
In the end, the rotor bearing spec and rotor balance tolerance/harmonnics may limit the gains and you end up with only a minor benefit of a 10% volume increase per main rotor rev. A similar effect could be had by simply increasing my crank pulley by 10%.

It has been a time consuming but I think a beneficial hunt to say the least.

I need these rotor dimensions....please someone!
............

wow, an utterly useless reply but I didn't even know what you were talking about without trusty google. Pls keep the post updated regardless. Cheers,

For comparison sakes in the ford world going from a 1.7L kenne bell twin screw to 2.2KB t/s will net nearly 100whp on 4.6L low comp/forged block.