A couple of PDF exploits from iMessage and Signal leading to code execution, and an app that roots the phone. Also, they seem to have managed to make their malware survive a reboot, which is difficult. That's another reason why automatic reboots are considered a security feature: exploits leaving malware in-memory are a lot easier to build than malware that survives a full reboot, passing all stages of signature checks and validation.

Yes, why don't they just say it instead of posting videos. Opening a sinister PDF in iOS can give them access to what exactly? anything/everything? What iOS versions or devices, etc?

Rooting, pretty much. Aka, executing code which shows their flying logo, but could also be exfiltrating data without displaying anything.

We don't know if Calif bypassed MIE.

When they hacked the M5 (one week after release!), they used a data-only attack, which is outside the scope of MIE. That could be the case here, or something else entirely. We don't know.

https://news.ycombinator.com/item?id=48139219

iOS does sandbox their processes really well, that's why this hack was as difficult as it was. The Safari thing you're talking about is JIT - iOS disables all runtime code generation across the system except in Safari (for JavaScript), but this is a completely separate thing from sandboxing.