📢 Three-transistor-class AB-Amp-modded-to-remove-DC-through-headphones 📢

I wonder, why is DC so bad? Is, say, 1V DC more strenuous on a speaker than 1V RMS AC? If so, why?

Well, this ain't a simple question. You see, 1 vDC on the old speaker, and let's imagine it's even wired up correctly, represents a permanent push on the voice coil. Effectively, the speaker is of course a highly sensitive solenoid. So, we then push the thing out and subsequently shovel our AC on top of this mechanism.

Then, we do get a heating effect, from the DC because it's permanent. There's no time for the voice coil to 'relax' and this generally is not a great time for the sensitive thing. On top of this, is the frequency characteristics of a voice coil that has been pushed out when it was very carefully designed to exist at a place allowing +- volt / mechanical movement. Effectively, you are infecting the speaker device with a disease, for which there will be no vaccination that's properly tested. During its disease period it both sounds like it has a bad cold and as yet $AstraNinja nor $Pfizer are actually tested properly. This in itself is bad science!

I’m kinda curious, What is the 100K Ohm resistor doing, it’s definitely important cause there seems to be almost no gain when you remove it…. I assume it provides some feedback, but why is it necessary?

It provides bias for the current source transistor. It also provides a corrective feedback to supplement the crossover compensation provided by the two signal diodes.