Since they have a speed limit you have to keep track of all gravity waves associated with all particles of the universe throughout all time and space.
So all particles still interact with all other particles, all the time, it's just that they do it with a time delay.
If there wasn't a speed limit it would be much simpler because all gravity interactions would be instantaneous and you wouldn't have to keep track of gravity waves.
However, since the speed of light is miniscule compared to the size of the universe you can ignore all but the most local interactions, and just schedule a computation sometime in the future when you know that the light vector will interact with something.
While instant calculations would perhaps make for a simpler system conceptually, the speed limit and locality principle ensures that less processing power is needed (at the cost of a lot of memory).
With a effective speed limit to space-time, you can "localize" the computation to the spaces where light has reached. And who knows, maybe light can't travel forever, it might just disappear after crossing some distance we still haven't measured (how we'd do that, who knows).
Giving yet another evidence to the "grand simulation" theory. "The universe" is just a group of simulated worlds connected by interacting photon particles (light).
But my point was that this also makes the universe more complex than an alternative fictional universe where information can be accessed instantaneously across any distance (which still allows for distributed computation, if synchronization or lazy computation is possible).
Some processes that are outside of scope we can sense seems like a too cheap explanation.
But we can deduce from various cues that it is being "simulated".
The Double slit experiment is one, experience of deja-vu another, dreams that partly manifest in reality after some time, the apparent speed limit of light, out of body experiences, the fact we are the only local top intelligent lifeform in this part of galaxy, etc...
All signs of processes and memory "bugging out". Except the last one, that one seems to be by design.
Where are mesoscale "bugs" of cells or the like?
Or is it only on the level of quantum effects and errors of the brain, wherein you're overlaying quantum jargon onto another iconprehensibly complex function?
Nothing, it was just the most obvious example (to me).