I find that about as convincing as "if you see a watch there must be a watchmaker" style arguments.

I don't see the similarity here.

There are a number of ways theorized to test if we're in various kinds of simulation and so far they've all turned up negative.

Oh?

String theory is famously bad at being usable to predict even mundane things even if it is elegant and "flat" is not the same as "infinite".

It basically makes no new testable predictions right now. Doesn't mean that it won't do so in the future. (I have no opinion about string theory myself, but a lot of physicists do see it as promising. Some don't. As far as I know, we currently know of no good alternative that's less weird.)