Or they'd need to find a way to break SHA256. That's a risk you don't hear much about.
I think it's fairly destructive to just discount new technology based on feeling. It really is a fun technology to learn and the more minds like we see on HN that research the topic, the more likelihood we will have in expanding the usefulness and security of blockchain use-cases.
Bitcoin promised to be a decentralized alternative to fiat currency and failed when it couldn’t scale to be even a tiny fraction of the global transaction. When people tried to improve on the original proposal they also failed at the very same thing while inventing new problems in the mean time. As of yet the only realized potentials of the technology are scams (which were never promised—at least to my knowledge).
Compared to Web 1.0 I think it is pretty safe for you to discount the technology based on a history of scams and failure to realize promised potentials.
If it's illegal to transact with the US under a previous version of this hypothetical blockchain banking technology, that's your 51% right there.
How could blockchain allow legal banking institutions to refuse to follow the law?
You only mentioned BTC but that’s just one coin. Several alt coins (with multi billion market caps) have had 51% attacks including ETC and BSV.
But anyway the original point was that it is literally wrong to say that blockchains are immutable. You didn’t address that point at all, so I guess you agree with it?
I mean, it's pretty easy why blockchains are considered immutable, you can even code one from scratch in Python in less than 200 lines of code.
To control 51% of all those hashing power, an entity would need to power computers that works as much as that, along with paying for the equipment and energy consumption and whatever costs needed for the operation. Current estimates of Bitcoin mining energy usage stands at 71 TeraWatthour, enough to power the whole country of Austria.