Sorry, I am not a teacher. I like to chat about electronics with guys who know electronics. I have helped thousands of people who had problems with an electronic circuit.
Do you know Google? I entered Attenuator (electronics) and found many descriptions and this schematic which is the same as your first one and the same as mine. The circuit you found with the attenuator feeding the LM3915 does not need an attenuator because its amplifier does not have much output power and the input of the LM3915 is fine with up to 35V.
I did not learn about the LM3915 in school. Instead I heard about it, looked at its datasheet (
www.datasheetarchive.com), bought one and used it. Its datasheet shows 3 peak detectors.
I went to school before opamps were invented so I learned about them myself. I also learned about many transistor circuits myself.
1) The Chinese amplifier probably does not have enough sensitivity so the opamp is needed to increase its sensitivity. I designed the opamp to be powered from a 12VDC supply. The datasheet and I said that peak levels would appear as a dim blur and that a peak detector circuit would make the LEDs look much better. I planned for the simple transistor peak detector would also use the 12VDC SUPPLY. But the simple peak detector I copied from the datasheet of the LM3915 needs an AC input that goes negative (the datasheet did not say so) and the opamp circuit with its +12V supply will not provide it, so I showed a peak detector using an opamp.
2) The Chinese amplifier has a lot of output power but then its voltage swing is so high that it needs to be attenuated (reduced). An attenuator is so simple that I thought you knew about it but you missed one of its two resistors. I showed the simple calculation.
3) Yes, each amplifier should have its own VU meter. It will show how close or far the level is from clipping, a clip/peak light does not. I never used a clip/peak light, instead I use my hearing.
You are very close to this project's completion.