OK, progress is being made....
It appears the signals are arriving at the amplifier correctly but the first stage (Q41) has wrong bias conditions. What you should have seen is 135 & 137 having similar levels and 147 & 149 having similar levels, maybe a little smaller on the 137 and 149 than at the inputs. I presume by 'half sine' you mean a sine wave going in one direction but clipped off on the other half, this indicates Q41 is not running in linear mode but acting more as a diode.
This might be tricky but using your testmeter again, can you try to measure the voltages between the base and emitter pins of all the transistors please. Ideally use the same probe on the base each time so I can check the voltage polarity.
Trying to explain simply, each stage of the amplifier adds a little amplification but in order for the output to be able to equally rise and fall in voltage so it can drive the loudspeaker, it has to start off at half supply voltage (positive ends of C416 & C417). This gives it most headroom to go up and down and hence produce most sound volume. Unfortunately all transistors and other components are slightly different so if the voltages at the first stage were set, by the time any errors were amplified they could be significant at the output. To combat the error, some of that half voltage at the output is fed back to the input stage through R406 & R407. If the voltage is too high, it reduces the bias on the first stage to bring it back until equilibrium is reached. We call it 'negative feedback'. What we don't want to do it feed the audio back as well because it would also try to hold the signal voltage at the output steady and that means no sound! That is where R411/C411 and R412/C412 come into play, they reduce the signal in the negative feedback path but as capacitors don't conduct DC, they still allow the steady voltages to stabilize. In fact those components set the overall gain of the whole amplifier.
Back to the fault, something is upsetting the DC conditions and causing the wrong feedback voltage to go back to the input stage. It makes a vicious circle, the fault is preventing the DC conditions from stabilizing. If you can let me know the B to E voltage on each transistor it should be possible to find where the break is in the feedback loop.
Brian.