Its good that the beams can be matched, but there are a few more considerations to take into account:

* How does this optimize the geometry to make the readout more insensitive to vibrations? Will it all fit inside a well insulated plastic box?

* How does the final beam size depend on the physical parameters (i.e. path length, lens Radius error, etc.)? i.e. we need a sensitivity analysis.

* The divergence angle of the beam at the detector determines, in part, how sensitive this setup is to backscatter from the RFPD. How to minimize this? How are the specular reflections from the PD face dumped?