The basic idea with a seed phrase or any kind of secret passphrase is that it requires a certain amount of brute force computation to find it. What you're doing sounds like the same thing but making it easier by introducing a bit of obscurity.
Since the early days, people have been running software against the blockchain to find money that is weakly protected and I can only assume that they're also looking for schemes like this. Now, there is a computational cost to matching every bit of data against any other with an algorithm like yours but if it's cheaper to do that then to brute force a key, then it just makes your money less safe.
To the problem you wish to solve, which is resiliency against natural disaster or coercion, there's already multisig which can be extended with as many keys as you like, each stored in separate places which are secure against these various adverse events.
It's better to distribute multisig keys to each of these places than depend on placing everything in plain sight except what is essentially another password that is vulnerable to those same initial concerns.
Your suggestion is that the safest solution we have as Bitcoiners to secure our keys is by using multisig? I can't accept that as the final answer to this problem. No chance.
The idea of distributing my keys among friends (and giving them veto power over my Bitcoin in a democratic process), or relying on a third party to host partial keys, is not a solution I agree with. Multisig wouldn't even help in this hypothetical act of God situation anyway, considering the likelihood of my friends being in the same disaster zone. It would probably only exacerbate the issue. And relying on a trusted third party service goes against the principles of Bitcoin, period, so that's a no-go as far as I'm concerned.
My goal is not just to solve the problem of losing both my hardware device and steel plate in a natural disaster. I want to address the issue of having to store keys on a physical medium altogether.
I'm wondering if there's a way to fragment, encrypt, inscribe, and eventually recall and decrypt the data in such a manner that we can feel secure in our ability to recover access to our accounts while rendering brute force attacks on the publicly stored info moot. This could enable the average person to feel more comfortable storing their wealth in a decentralized, trustless, non-state-affiliated digital protocol like Bitcoin, instead of relying on centralized banks and fiat currencies.
Key storage is obviously a hindrance to adoption. This is a digital currency, a "digital revolution", and here we are engraving words into fireproof steel like blacksmiths in medieval times, and if/when we lose it or someone finds it, it's game over. You could multisig but the more keys you have, the more keys you need, the more keys you need to store.
I came across https://www.borderwallets.com/docs/the-problem, which explores some new innovative ways to secure keys without relying solely on word memorization.
While multisig serves specific use cases, I still believe we are in the infancy period of key storage and "account" recovery, despite us going on almost 15 years of block discoveries. We need to push forward and continue exploring new possibilities to make key management, storage, recovery more user-friendly for the masses and even just more aligned with the digital nature of Bitcoin itself.
reply