No. I've had the same feeling with some of his students. I can even remember collegues of mine walking away during talks, because it was one hour of handwaving speculation.

Sometimes that can lead to nice observations. But let's not overdo it :P

You probably know this, but it's worth clarifying that you do not need interactions for entanglement. Even a free quantum field theory has an entangled vacuum state.

Yes, but I actually wanted to say that what we call "free" field is actually an interacting field. Namely, the term ##(\nabla\phi)^2## in the "free" field Lagrangian density is actually an interaction in the same sense in which harmonic oscillator potential proportional to ##x^2## is an interaction. Indeed, you probably know that "free" field can be thought of as a series of an infinite number of coupled harmonic oscillators.

A concept or procedure to produce a number during a measurement is observer dependent if different observers follow the same "prescriptions" from their perspective and later compare their numbers they disagree.

A typical constructing principle for physical law is that laws of physics - to the extent that they are objective - must be seen the same to all observers.

When apparent disagreements appear, objectivity are often restored once we find the transformations that relate observers views and the laws can be formulated in the objective way. That was the case with sr and gr as well. But the observer dependence ive talked about relate observers that cant be generated just by simple diffeomorphisms. There is no known symmetry for this.