That's what the app sandbox is for. As long as the sandbox holds up, it doesn't matter what security vulnerabilities are in the app within. And if the sandbox doesn't hold up, you've probably lost the game regardless. There's plenty of room for security vulnerabilities in an app like Chrome even if they're not using a custom browser engine.
App sandbox is just one layer of the security model that iOS employs, and another is not allowing JIT'd code, in case someone finds a way to drop a payload into W&X memory. There's no reason to increase your attack surface if you don't need to.
Is the restriction on JITs really a security thing? That seems so completely far out there on the list of concerns compared to common stuff like buffer overflows which Apple does pretty much nothing to prevent. It always looked to me to be a technological means of enforcing their rule that you weren't allowed to download new code to run (on purpose).
It's very much a security thing, there've been lots of exploits that work by dropping a payload into a W&X marked area in browsers (usually dropped in by buffer overflows).
Apple also doesn't like unsigned code, and JIT (or self modifying code, or whatever falls into that category) is inherently unsigned code.
