Most of the circuits that work well use the photo-transistor with gain so its output is from its collector and its collector resistor value is high. But then it needs to drive an amplifier with a high input impedance like the input of an opamp. The opamp can also have as much gain as you want. Then the opamp drives an LM386 power amplifier that drives a speaker or headphones.
The last circuit you posted has the photo-transistor collector resistor value only 1k to 5.7k (instead of 100k to 1M) and it feeds the fairly low 10k resistance of the volume control instead of feeding the many meg-ohms input impedance of an opamp.
The circuit is also missing the important 10 ohms resistor in series with a 47nF capacitor to ground at the output of the LM386 and is also missing the very important 100uF and 0.1uF capacitors across the battery.
If a transistor replaces the opamp then the low input impedance of the transistor kills most of the output from the photo-transistor, which is why one of your circuits used the photo-transistor as an emitter-follower that has a low output impedance but has no gain.
What are you using to provide a modulated light source? Maybe it is not modulated enough.