Research question: What is the biggest unsolved security gap in self-custody under physical coercion?

I'm researching a specific question about crypto security:

If someone physically coerces the legitimate owner into authorizing a transaction, what protection does the owner actually have?

I'm less interested in theoretical attacks and more interested in how people with meaningful crypto holdings actually structure their security today.

For example:

• Multisig
• Hardware wallets
• Passphrases / hidden wallets
• Duress or decoy wallets
• Timelocks
• Spending limits
• Geographically separated keys
• Trusted third parties / guardians
• Keeping only a small amount readily accessible

My main research question is:

What is the biggest weakness or failure mode of the approaches above when the attacker has physical control of the owner?

And a second question:

Is there any setup today that makes it technically impossible for a coerced owner to immediately move their long-term holdings to a new address?

I'm interested in real-world setups and experiences, not product recommendations.

If you've thought seriously about this threat model, I'd be especially interested in what you currently do — and what you still don't feel comfortable with.