Apologies if the following point has already been discussed ad infinitum elsewhere, but I've noticed that whenever I have sent BTC, the transaction has always been published instantly, even though it took a few minutes to actually confirm. If the Bank of America or Visa were to use the BTClockchain for their own internal purposes, they would certainly have procedures in place to prevent their employees from trying to double-spend the BTC sent to addresses to establish transactions. Since such published transactions are sure to be confirmed, shouldn't publication alone be sufficient both for legally accountable record-keeping and to initiate or complete smart contracts? If so, doesn't this mean that, for such uses by such institutions, there is no transaction-time limitation?
imagine it as if funds where debited out of your account instantly. but the recipient cant use them yet as they havnt been credited to him until its confirmed.
many people can still mess around and cause issues before its confirmed. and so for large purchases its not enough to rely on the publishing of a tx, but rather wait till its confirmed.
bitcoin is still not perfect in regards to the limbo between publishing a tx and having it confirmed.. however there are many idea's in the woodwork that are all aiming to sort that out.
that way payment systems can have 100% trust that once a transaction is published, its then locked in the waiting list to be confirmed. and any attempt to respend the same origin funds elsewhere will be ignored