Hi Tim,
I think this is a great project design. One thing you could do to make sure your design is robust is simulate the effect of self phase modulation (and XPM) on the signals. I noticed that the output power was around 6 mW, which could be large enough to make a difference on the BER. SPM/XPM could have an even more important effect since you are using dispersion compensation in the design.
Try replacing the Raman Amplifer fiber with the Bidirectional Optical Fiber component (there is an option in the component properties for enabling Raman scattering). You might be sacrificing simulation speed, but the model also includes SPM/XPM, which will increase the accuracy!
Cheers,
Damian
.
