> Making a transaction this large is a security risk. It effectively establishes a billion dollar bounty for any party who can rewrite enough blocks to erase it
... accomplishing exactly nothing as all that would do would be returning the money to original spenders. Really, the amount of misinformation floating here is unexpected.
> Now that the value resides in a single coin, any subsequent transaction faces the same risk.
Non sequitur. A rollback of such a block would, by significance be somewhere between "mildly annoying" and "introducing suspicion that future transactions could also be rolled back." In any case, coins would end up where they are supposed to, and no theft can be done. There would be increased risk of double spends, but worried parties could always invest in more (legit) hashpower.
> A consortium/trust has formed in which individuals pay into a common pool of money.
Speculation, but reasonable. It makes business sense.
> That money is then protected with a multi signature script (consistent with the P2SH type). Given a threshold of signatures, the money can be spent, subject to other constraints. These will remain unknown until the first payment is made. At that point we'll know the number and identity of all the eligible keys and the threshold needed to make payment.
Speculation, but too wild. Basically wishful thinking.
> Edit: to be clear, the "bounty" I'm talking about can't be directly claimed just by mining some blocks. Instead, it would have to be claimed as part of a double spend of either the transaction in question or as subsequent transaction of the now-enormous output. Most likely, there would be collusion of some kind between a miner and the owner of the keys. I'm not saying this will happen, but the bigger the transaction, the greater the risk.
Again, wild speculation. So the theory is that someone collected $1B in BTC just to pull off a massive double spend stunt? At that level, it's likely that rubber hose cryptanalysis would resolve the issue quite efficiently.