I've assembled a prototype on a breadboard as a sanity check and also for testing passive filtering. I still had hope that I can cut enough higher harmonics on each string so that the main frequency would have the highest amplitude of all (which is not always the case with normal string plucking).

Well, the tests showed what I could've guessed beforehand: the highest playable note on each string is as high as two octaves above (so, fourth harmonic) the lowest playable note; therefore, there's no way to filter harmonics without dramatically dampening all of the notes on the higher frets. Well, quick and dirty MIDI tracking is out of scope now for sure.

But at least I see that the amplifier works as intended and can transfer the content of the breadboard to a first prototype PCB.