I apologize for going meta on this, but I literally cannot make heads or tails of his writing, which I'm pretty sure is about something that I'm interested. With a name like "Andrew Cooke" I'm inclined to think English is his first language, but...wow. Is this a style thing? Can someone summarize?
logitech made hardware and software that allowed you to stream music from a linux computer to elsewhere in the house. they have stopped making the hardware and the software appears to be "decaying".
forking and fixing the software sounds like the obvious short-term solution, but it's not so easy - it's fairly opaque perl, and despite various interested people and an open bug report, nothing has happened so far. also, it's not a good long-term solution, since the hardware will fail eventually, too (the hardware was pretty dumb; a lot is done by the software/firmware).
the best alternative i have found is the audio engine d2. that works at a lower level than the existing logitech devices - it looks like a USB sound card to the computer. that means that it works with a variety of different music players (avoiding the trap i fell into with logitech's software; on the other hand, that also makes it less flexible, so the kind of setup buro9 describes above is harder).
does that help?
one piece of context that may be missing is exactly what this hardware does / did. typically in a "computer audiophile" setup music starts in a digital file on a disk, is sent somewhere, converted from digitial to analogue, amplified, and fed to a speaker. what logitech did was the "send somewhere" part (plus, optionally, conversion to analogue). so you could keep your music in one place, but listen to things elsewhere (or in multiple places - maybe your main speakers in the living room; a headphone amp in the bedroom; monitors in the office). effectively it was a "wireless digital cable" (plus software router).
it was also useful that the logitech devices could work with a purely digital signal. that let people use other (typically more expensive, better sounding) hardware to do the digital to analogue conversion.
I'm assuming that since you have the ability to fork the software that you know the API for the device. Great, then it can be rewritten in a language other than obfuscated Perl.
Beyond that, if it's as dumb as you say, how about a Kickstarter for a new hardware output endpoint that can respond to this new, rewritten software? Perhaps an Arduino and DAC can be hooked up satisfactorily in the meantime.