It's not that kind of a hologram... all the hologram idea says is that all of the information in our universe can be represented on a 2 dimensional surface. It doesn't claim there is literally a hologram projector somewhere projecting a hologram of our universe. Such a thing would be nonsensical. If the projector were outside our universe it would, by definition, be causally disconnected from our universe and thereby never capable of influencing it in any way. If it were capable, it would, again by definition, be part of our holographic universe.

Just because something is outside something else does not mean it is causally disconnected.

The computer program runs inside the computer. The computer is outside the program. But nevertheless the computer's hardware is causally connected to the program. Further still, the programmer is outside the program, but he very much can influence it. Indeed, he brought it into being.

The idea is that to be outside the universe is to be causally disconnected from it, not to be outside of some arbitrary system. A "universe" is not an arbitrary system -- it is the whole of reality.

A computer program is not a universe independent of the programmer. They both exist in the same universe because they are causally connected. If they were in different universes, they must be causally disconnected. You're whole argument rests on the assumption that the program and the programmer reside in different universes, which is clearly not the case.

Field theories like QED have the concept of a field, which is something that permeates all of space, which makes it non-local. Electrons all have the same electric charge because they are all different excitations of the same electric field. Does that in some way mean they are all the same particle? They are all connected by the same electric field and have the same charge.

You can reproduce entanglement in a number of local theories. I always mention Einselection in these sorts of threads, because it's a very clean way to get something that looks like many-worlds but isn't hand-wavy and can reproduce entanglement via local mechanisms. In particular, contradictory measurements may take place, but they are forbidden from interacting via superselection rules (such that <x|A|y> = 0 for any observable A for contradictory states x and y). The math isn't as developed as Pilot Wave theory, but it's at least as aesthetically pleasing (deterministic, local, etc.). I guess it gives up definiteness, but in a less objectionable way than many worlds.

Are you implying that Many-Worlds is hand-wavy? It's not.

A large fraction of working physicists subscribe to the Many-Worlds Interpretation, which preserves locality even though it has entanglement.