Yes.
Probably, but until I dig them out of eight year old archives, you'll have to assume that my recollection isn't failing me, if only for the sake of argument.
It was possible to manage this, real time, back when we had single-core CPUs, and a few hundred mebibytes of system memory, when broadband was just catching on. If there are insurmountable technical barriers to tracking and depicting three figures of players in SC, that's a pretty severe and seemingly inexplicable downgrade for anyone who has played these other games.
Identifier, state/positional data, and movement vector, times tick-rate. That's a couple KiB per second, per entity; less if the latency compensation is good and tickrate falls off with distance. I've heard that SC had issues with this years ago because it was always tracking everything (individual characters and every piece of cargo within a ship) in detail at all times. I'm not sure how much of that is true, but that itself sounds like a problem. None of that should be relevant until extremely close range and even then only if separated from the rest. Abstraction and simplification should be the rule anywhere the detail isn't actually required. Surely the renderer is doing this for graphics (via LODs).
So yeah, I'm not well educated on these details with regard to Star Citizen, but that doesn't change my sense that if it was possible for other games, it should be possible for a much more recent and advanced game, especially one that's received this sort of backing and seen this much development time. It's like EDO's performance vs. Horizons...I'm sure there is an explanation why EDO is ~40% slower while rendering a visually indistinguishable scene in a completely GPU limited scenario, but understanding that reason isn't going to excuse it.
