This topic may deserve a bit further discussion, now that it's clear that buying and selling examples is 'off the table' here.
I recall being fascinated by this in the mid-Nineties... about the time the NMRA DCC standards were being finalized. It wasn't the first system to 'inject' an AF or RF (I believe the MRC device used 400 KHz FM modulated) signal onto the track to feed audio to an onboard speaker, but it was notable in allowing a user to play 'audio' that could be of considerable complexity, or programmed to play extended clips sequenced externally, to what sounded like "the moving train". (Didn't the "Rabbit" wired TV extender use this kind of low-RF FM modulation in the '80s?)
I suspect if you didn't keep the track scrupulously clean (or, ideally, gleamed it and then kept it in shape) you'd have dropouts or noise in the sound.
As I understood the literature, you could have larger 16-ohm speakers than the stationary ones to be mounted in buildings or under the layout, to get better than transistor-radio audio quality. My thought at the time was that the 'first best use' of the stationary output was to impedance-match and attenuate it to a typical consumer-audio line-level input (there were cheap modules, then as now, for connecting things like car radios to larger amplifier-speaker sets without a dedicated preamp) and it would not be a vast leap of understanding to get from there to a Rolling-Thunder-style subwoofer to synthesize relatively nondirectional low frequencies for the stationary rig while retaining higher-frequency signals on the 1" onboard arrangement. This might have gotten over the power issues that kept it strictly 'either speaker A or speaker B but not both'...
I think the NMRA's modulated-signal-over-power approach to DCC probably invalidates any system that puts a modulated AF or RF signal across the rails, so this is not only a strict DC solution, but one that might not tolerate high-frequency PWM that can't be filtered effectively from the vicinity of the 400 kHz carrier. (I'm pretty sure the choice of audio carrier was made to prevent a wide range of PWM modulation schemes and frequencies from affecting the audio quality or signal characteristics...)
I believe we've had a number of discussions about this idea in practice.