I also wish they'd have started off like that from the beginning.
It might have taken them a bit longer to get the initial implementation out the door, but it would have avoided all of these discussions, which I'm certain are also impacting WebAuthN/passkey adoption across the web (assuming that the views on passkeys here are at least somewhat representative of technical decisionmakers).
But if the choice is between getting to the status quo via suboptimal intermediate steps or not having WebAuthN and passkeys at all, the former seems much preferable.
It might have taken them a bit longer to get the initial implementation out the door, but it would have avoided all of these discussions, which I'm certain are also impacting WebAuthN/passkey adoption across the web (assuming that the views on passkeys here are at least somewhat representative of technical decisionmakers).
But if the choice is between getting to the status quo via suboptimal intermediate steps or not having WebAuthN and passkeys at all, the former seems much preferable.