Realistically, if your threat model is government secret services, and you're using unmodified consumer grade electronics, then you're in 'danger' no matter what. You can't effectively mitigate a threat at the state level using resources produced under the watchful eye of the same state. All they have to do is ask the producer to swap out the device they gave you with a device that comes compromised out of the box. And that's assuming the tech is perfect. Most likely they just hire someone to defeat the countermeasures. However many resources Apple has, I assure you even the most janky state has more.
Commenter above was saying though that the device's anti-tamper tech would save you from state level attacks. I'm just getting at the fact that that's not going to work, since if a proverbial 'they' want to take you out, there's other ways to do so you can't overcome. Just a few examples that came to me about how easy it is to foil anti-tampering measures.
> It seems obvious to me that if you're being targeted, tailed and tracked and probed, you've already lost.
Depends on which government agency watchlist you are. If you are some sort of Islamist terrorist, the tools that are open to the government are far more capable than if you are some sort of low level drug dealer.
I find it ridiculous that people build threat models around organisations with almost unlimited resources that will only care about you (enough to tamper with your hardware) if you have done something very, very wrong.