Didn't care, cloud storage?
Yes, it all started from this huge, non-standard, almost 4 MB block. If the main goal for Ordinals would be something different, then they would start from something else. Not to mention that if you want to test edge cases like that, then it is way easier to do so on testnet3, signet, or even your own regtest, than to ask some pool like Luxor. And because there are a lot of non-Ordinals cloud-storage cases on testnet3, then they could focus on encouraging those users to use Ordinals first, or even just trace their "non-Ordinals inscriptions" (also because testnet3 allows non-standard scripts, so you can do more things with Ordinals on testnet3 than on mainnet).
But aren't all sats in your wallet, which can only be moved if you control the keys, already represent actual ownership of the "property".
Ordinals don't use those properties. A lot of wallets can change the order of your inputs or outputs, and boom, it breaks Ordinals. Also, because Ordinals don't care about sighashes, they are wide open if you don't use SIGHASH_ALL.
My personal opinon. If there's anything to like about Ordinals "as a project", it's because they didn't ask for permission to start building. They just built it, just like how the Lightning Network developers built their network.
There is a difference between "permissionless" and "testless". Ordinals were "testless", because they didn't write many tests they should, before going into mainnet. And that's why ownership is not enforced (nobody tested that), that's why zero satoshis can break things (nobody tested that), and that's why sighashes are not used (nobody tested that). Not to mention things like CoinJoin (also nobody tested that). Also, there are probably even more bugs in mimetypes or recursion, but they will explode later, because it is the same situation (nobody tested that).
In case of Lightning Network, it was tested, at least partially, before touching mainnet.