From my understanding (though no deeper than what's in the Bitcoin whitepaper), the central new idea (and it is indeed a significant contribution to CS in general) is laying out of how to build a distributed timestamp server, and then distributed ledger from only untrusted peers by using existing technologies circa 2009 such as p2p, hashing data blocks, priv/public key for verify/sign, etc.
For instance, people in auditing and anti-subversion fields promoted schemes that allowed multiple, untrusted parties to accept a result by one party doing something whose logs were checked by others, partly (esp randomly) or totally. The scheme required nothing but a database or filesystem on inexpensive hardware for most of it. Only crypto was hash of some set of transactions or big file with a signature on it. Soooo efficient. Also builds on stuff we have decades of work on securing with full, battle-tested stacks for internal components.
And then there was Bitcoin with its long transaction times and insane work in mining. And use of begging to avoid a key risk at one point. That's innovative as it's the first time I've seen it in decentralized security.
That's pretty much how genius works much of the time, seeing possibilities that exist with existing stuff that haven't been done yet.
Doesn't make me feel dumber when I realize how few programmers or systems people are re-using or recombining the best stuff from CompSci or past commercial activities. Versus how many ignore it to repeat same crap and failures. :)
> I would argue it represents a key development in the history of computing & networks, and possibly a new subdomain of Computer Science
Those two statements are in conflict. Anyone who's changed the history of computing and created a new subdomain of computer science by doing something no other modern cryptographer thought to do is a fucking genius.