Frankly even the idea that we could be in a simulation stretches the whole idea of what a simulation is to the breaking point---a simulation by definition isn't real, and yet here we are with actual conscious beings, and he's claiming we're simulated.

But we don't have privileged, direct access to the real world anyway; everything you experience now, is, in a certain sense, a "simulation" constructed by your brain. (If you don't like the word simulation, you're welcome to choose another.) When you look at a red book, the reason you think there's a red book out there in the real world is because light reflecting off the book is being absorbed by your eyes and translated into sensory data that is sent to your brain. If we replaced your eyes with some as-yet-science-fictional camera that supplied the exact same data to your optic nerve, you might not notice; you don't have any reason to care whether the information from which your visual field is constructed was gathered by a "real" eye or a merely artificial camera. But then if we put a shutter cap on the camera and started supplying your optic nerve with data that was generated by a computer program rather than by means of measuring light, you again have no particular reason to notice or care. The hypothesis "I'm experiencing the real world" and the hypothesis "I'm being supplied with real-world-like sensory data despite being implemented in some other way" make the same predictions. We might have any number of good reasons to reject the latter hypothesis, but "simulations aren't real by definition" isn't one of them.

his whole equation relies on the ridiculous assumption that the number of individuals in a simulation is equal to the number of individuals in a real universe (he calls both H)

One would imagine that assumption was made only to simplify the presentation; it doesn't affect the core ideas. For example, see Robin Hanson's "I'm a Sim, or You Aren't" for a variation that makes different assumptions about the size of simulations.