I get that direct evidence of a leak is difficult. However, a sudden surge of master passwords being known by third parties in uncorrelated accounts is a very good evidence that something happened. If that's not what happened, then what happened exactly? Was it really a bug in the notification system? Do they have evidence that the password used in the blocked login attempts weren't really the actual master password?
There is a lot of things they can do to show they are on top of things.
I also agree with you about the 100% confidence about not being compromised. Perhaps my previous statement was too black/white. I don't want PR or placating statements, I want a transparent status report without weasel words and which exhaustively covered the different cases (e.g. SOME of the messages were sent in error. what about the rest? Are the rest routine compromises that happen normally? Or was there a spike in compromised accounts?)
They have been extremely clear that they have not found any signs of compromise. Did you miss that?
Of course no company can technically guarantee that they have not being compromised. If you are looking for someone telling you at any point they are 100% confident no user accounts have been compromised, then you will pick a company lying to you.