That's fine. These will be out of order from your original reply, but it was easier this way.
The 51% requirement is actually a lot *worse* than the supposed 0.000001% requirement of Zuckerberg. How do you think we got to that situation to begin with?
First of all, voters voted for politicians who didn't want Zuckerberg regulated, because spreading fake news was good for them. Facebook and others *could* be regulated, but US politics effectively prevents it. Voters - whether on blockchains or in real life - generally are not well informed, and most crypto users are catastrophically poorly informed. Donald Trump could probably get millions of Bitcoin users to download a 'Patriot app' and mine for him unwittingly if he wanted to.
Secondly, unlike blockchain, you can actually do something about Zuckerberg; that's what central regulation is. I also point out that 51% of Blockchain is not some nice democratic system; that 51% is quite easily going to be held by a small handful of big players, making it the same as the current system except without any possibility of stopping it.
Thirdly, if you think a wild west scenario can *ever* be a strong positive, you need to go back to history class. There's a reason we have literally thousands and thousands of laws about everything from food safety to business and everything in between, because a "no rules" situation leads to worse than disaster every single time.
"There's also the ability for you to choose nodes as verified/unverified and fork the network. Meaning if we know of one fraudulent transaction in a 51% take over, we can literally just ignore the transaction and hard fork the blockchain. Obviously they can't trade with us after that, but it sounds like that would be a positive outcome." <- no, this is a lot worse than you think it is, because you are assuming you *know* the 51% is fraudulent, which you can't know - there is no possible way to know.
"Outside of the blockchain, I think if you can convince the top 51% of any financial system to coordinately do something, theres not much the remaining 49% can do short of war. I don't really see the government as a guardian against that kind of "only top 51%", I mean Princeton kinda recently showed it to be the case" - There's a lot you can do, because 51% of the financial system is not equal to 51% of democratic voters (unfortunately, it *is* equal in blockchain, making blockchain extremely vulnerable). Your example specifically goes into American politics, which is very different since American democracy is a running joke in most nations - it's so poorly designed that most of the time there is minority rule there. Other democratic countries are faring far better.
"Agreed, its a shame this is also true for web2. Phishing attacks are hard to stop. Really its a good thing browsers have strong cross origin policies to prevent web2/web3 account data being copied too." - actually no, it's orders of magnitude easier on web3 than web2; there are no central authorities to declare a website genuine or not genuine, as you currently have. Web2 does an astronomically better job, largely because it does the job at all.
""client-server model throughout web history: because without it, verification in any real sense no longer exists"
What are your thoughts on private/public key verification? Like gpg keys" - they are worthless in terms of doing anything for Web3. They are used for preventing a communication being tampered with - NOT for verifying if a communication is valid. If Alice sends Bob a message using gpg keys, but in fact I stole Alice's computer and sent the message myself, Bob is not going to be aware of that. You can't fix the lack of central authority with encryption, encryption doesn't perform that function irrespective of what kind of encryption it is.
"And I recently learned more about DAO's which are very much a voting-type system where votes can be bought. DAO's seem kind of dumb to me too, I'm not sure what their purpose is outside of hype." - Votes are bought in any blockchain system, DAOs aren't relevant there. That's how any decentralised system functions by definition.
"Agreed. If we removed streaming protocols like WebRTC, FTL, etc and other non-file stuff from the internet, I think a web of just transferring HTML/JS/CSS image/video/audio etc files could be quite nice. Obviously it would be better with the additional protocols but still I'm glad were on the same page that peer to peer would work great for the basics of the internet." - Streaming is vastly superior to peer to peer for those functions for about fifty reasons, of which stability, predictability, geography, scalability and privacy are but five. We are not on the same page at all there. Even for basic things like HTML files peer to peer is vastly inferior; peer to peer is only viable when you don't care how long a transfer takes, whether it even completes, whether it's always available... the list goes on.
"In my university's distributed cloud computing class when we were going over blockchains, the professor discussed some experimental papers showing how a bad actor could use network interference to reduce the required percentage for a complete take-over to possibly 34%." - not realistically, no. It's plausible to take over *temporarily* in that manner, but only as long as the network interference persists; academically, it's possible that you could e.g. validate a fraudulent transaction in this way, but you couldn't permanently "take over" the network with 34%; that isn't physically possible. It would also be fraught with huge risks for the bad actor, who would likely find buying the extra 17% for a proper takeover much more lucrative.
