Well, at the weekend I finished building the unit and started testing it. The TX part worked first time, but the RX did not. Time to find out why!
Using logic and common sense, I carefully went through the various stages (8V regulator, later audio stages etc) and tracked the fault down to an intermittent preset SMA pot that sets the FET bias. The FET stage is the very high impedance stage that follows the PIN photodiode.
The error was entirely my fault and easily fixed by removing the part and redoing the surface mount joints. Now the full transceiver is working well (but yet to be put into 100mm optics) and the RX sensitivity is close to that with my K3PGP design RX. In the coming weeks I hope to get a transition piece to connect the transceiver "tube" to a 110mm drainpipe that houses the 100mm lens. I'll then assemble this onto a stable tripod with sighting scope and I'll be ready to look for QSOs.