Does it bear any resemblance to the parts you've posted?.
If so, it has far too many audio stages, almost none of which are required, and almost none of which have supply decoupling - these are adding instability, noise, and distortion - and nothing of any use. The stages appear to be thrown together, with no concept of design or calculating values - and with random darlington pairs appearing for no good reason. There also appears to be no sensible audio filtering, just odd capacitors here and there.
Likewise the rest of the stages are poorly, or not at all, decoupled (and mostly badly designed).
All you need for the audio stages is:
Detector.
One amplifier stage (optional).
Bandpass filter (simple opamp would be handy).
Power amplifier.
Just checked the two ARRL designs I mentioned earlier, one use a passive low-pass filter followed by a 741, driving headphones. The second uses a passive low-pass filter directly driving a power amp IC to a speaker.
Perhaps you would care to go through the design of one of the stages?, it's reasons for been there, and how (and why) you calculated the values.
Sorry, but it's truely the worst radio design I've ever seen, even beating the super-regen fisaco's of a year or two ago.