2. Is also a reasonable point to make, but the nature of the goals here means it has to do a lot more than just sit on top of the existing protocols. It's a secure, decentralised, autonomous network/platform, which needs to be built from the ground up in order to solve the many problems we have with the existing web.
I'm not sure what you mean by monolithic. The ideas are new and can take a while to grasp but it's worth it.
3. The community have had long discussions about this so many will agree with you. But the best response I've heard to this point was from Tim Berners-Lee when he and David Irvine were at the Decentralised Web Summit in 2018. Someone said something similar to your point 3 to David and Tim jumped in and answered that it wasn't necessary by saying: and what was the killer app when I invented the web? Or something along those lines.
I'm not dismissing your point and many of us aren trying to come up with ideas for such killer apps.
Personally I suspect we already have some - take a look at the SAFE fundamentals for example. But it may well be something nobody anticipated. It's fun thinking them up though :-)