To eliminate the risk of the trezor generating a private key I don't own, I'm gonna put its key on the raspberry pi zero and see that it generates the same master pubkey and shown in trezor. This proves I own this key, but it might be a kew that someone already owns. No problem, that's why I'm doing multisig.
Why don't you just generate private key or BIP39 seed words offline and then import them in your hardware wallet if you want?
Entropy generation in Trezor is not so great, and it is mixing external entropy from computer with internal entropy from built-in hardware random number generator.
Other hardware wallets like ColdCard are using secure elements for seed generation but I would still use old methods with offline BIP39 if done correctly.
I'll then receive some Bitcoin on one address, erase both wallets, restore them with the private keys, and then try to spend this Bitcoin, just to make sure I really owned the coins.
You don't need to first receive coins on your address, but you should first delete and restore your wallets before sending or receiving any coins.
That way you will be sure you wrote everything correctly and made good backup.
Note that Multisig for most hardware wallets are not totally secure and for Trezor wallet you can't automatically verify that the co-signers are safe, and it has to be done manually for every transaction.
Jameson Lopp did a test of hardware wallets with multisig few months ago and it interesting to read and notice many flaws:
https://blog.keys.casa/bitcoin-multisig-hardware-signing-performance/