BTW there is no DC offset OR bias adjust. And it has NFB tone controls (they're PART of the power amp's feedback loop).
There are 3 caps per channel after the volume controls/ mono etc stuff, feeding the power amp. Since all frequencies are affected about the same, we can discount the tone control's effect.
4.7uF feeds a bipolar transistor(2sc1919) buffer amp, then there is a jfet(2sk117) muting circuit that operates on power up, and during function switching. A 2.2uF cap feeds the muting jfet and a 0.47uF cap takes the output of the muting curcuit and feeds it to the power amp. Those caps could be shifting in value (reducing the volume) or leaking ( upsetting the biasing somewhere).
C313 and C314 (100uF/35v) do power supply filtering for those 2sc1919 transistors, and if they are leaky they could drop the voltages to those transistors, loosing gain. Check the voltages on both caps.
IF you had a constant amplitude signal you could feed in (test tone, interstation FM hiss etc) you could take ac voltage readings and see if you can spot a drop in the signal level at both sides of each coupling cap.