The idea of this reality being a simulation is very enticing in many ways but it lacks a firm way of us proving/disproving the theory. Almost every scientific approach to the problem could be dismissed as being manipulated by the creators of the simulation. It doesn’t mean that it isn’t worth thinking about but it may ultimately prove very frustrating to prove/disprove.
There is one note of caution as we actually start testing ideas. We have no idea about the programming skills of our coder overlords. From years of programming myself I would be a bit worried about their error handling. Are we the result of many iterations of legacy code jumbles and half baked integrations?
If we poke too hard do we cause a problem in their code? How do they handle that? Is it robust or do they just save the event to a log file and reboot with a fresh code base? Hmmmmm, I wonder how many times that has happened already?