js: On the Luxman it was to be expected anyway that it wouldn't work, considering the existing controls - and the block diagram confirms it. And what will happen with rec-2-out, when rec-2 is selected on the record-out selector, is that the controlling microprocessor will disconnect rec-2-out, 'cause otherwise rec-2-out and rec-2-in would be hooked together.
The question really is the exact implementation on the Accuphase: From the chosen approach and the existing controls it would seem very well possible, that it would work - but without further info material (or the chance to try it) one can't really be sure, that it actually does. Hence my two examples above (both service manuals should be available for download over on HFE): On both models the chosen signal routing approach and existing controls would be similar to those on the C-2000, but if you check the block diagram of the hk825, you can see why it would work on that one. Whereas in you check the block diagram of the PM8000, you can see why it wouldn't work on that one - and rather unnecessarily so at that, 'cause with a slightly different wiring of the record-out/tape-copy selector it would work...
So the question would be, whether the C-2000 does it like the hk825 - or rather like the PM8000, which unnecessarily disconnects the output to the delivering tape-loop from the selected source, as already mentioned above.
tom3: Well, the bonus of an approach like on the hk825 is that the equaliser will require no other extras than just an input and an output, because you can already control everything from the pre-amp(-section), having the full choice between listening & recording via EQ, listening via EQ & recording flat, listening flat & recording via EQ and listening & recording flat (at the same time each). Whereas if one would connect the recorder simply in series to the EQ in the same tape loop, listening via EQ and recording flat wouldn't be possible.
Greetings from Munich!
Manfred / lini