Soundness of Lean requires more than correctness of the kernel - it requires that the theory be sound. That, frankly, is a matter of mathematics folklore. "It is known" that the combination of rules Lean uses is sound... unless it isn't.
The "winning" offer was formulated such that its cash value depended upon the next best offer. I.e., when you strip away the complexity it was of the form "next best offer + $". That this bid "won" indicates the "auction" was not a sealed-bid auction of the kind you are thinking of.
Wait no none of this is right, it doesn't make it an unsealed auction just because one or more of the bids has terms and conditions. This is how every real-estate transaction is conducted.
What makes it fine in the end is that once you collect all the bids and resolve the legal paperwork you can assign values to the individual bids and compare them which is how people who sell their homes choose what offer to go
with. And home sellers do consider intangibles and go
with lower-cash offers all the time.
I don't really understand what all the legal kerfuffle is about. Is the court really going to force the families into a sale they don't want when the sale is for their benefit however they choose to define it?
> It was specifically structured to give better-than-next-offer
That is the most farcical bit. When you strip away all the legalese, the offer was literally formulated as "next best offer + $". That is not a valid sealed-bid offer.
It might be missing a few logging roads, private driveways and carpark lanes - but it shows Google is entirely capable of mapping streets in detail all over the world.
And in those same cases, you could probably safely do the same with FSD (s), but because FSD isn't as limited and the Mercedes product is not a serious competitor, there's no compelling reason for allowing it.