I feel like this assumes simulations end at scripted events and go no further.

"A simulation simulating an apple is not equivalent to a simulated apple." is my retaliation. Our simulated apples do not need to feed bookworms or robins - but we could simulate apples which feed simulated bookworms. In fact, the game of "Snake" is greatly similar to that concept! The argument that we live in a simulation simply means there are rules where we cannot eat some things and can eat other things. In an OO world they might have "isEdible: true". A simulated apple is edible. A simulated simulated apple is not. No inheritance!

There is an upper threshold for how good graphics can become: "realistic". I don't mean 90's realistic, I mean unrecognizable from what we call reality realistic. On small scales (not simulating many things) we are already nearly there. What comes next is more computational power to scale that level of detail to entire worlds instead of single objects. The argument against some 90's flight sims graphics being hailed as "realistic" back then doesn't mean the bar for "realistic" changes. It's using a different definition of "realistic".

There is no way to prove or disprove we are living in a simulation. Any philosopher would know you cannot know what is outside the limitations of our universe. The more convincing argument is that there is a lack of evidence, as with any other belief that requires knowledge of "outside of our universe". You need to be able to look outside the box and claims about the outside cannot be made from inside the box. There is a very real possibility that we are living in a simulation - there is currently no reason to believe in such a claim.

When we successfully run our own simulation of "everything in our universe" the idea might become more plausible and tasteful, rather than extrapolating from our current technology that some society may have been able to and did so.

That's what 98% of the simulations say.