Being someone with first hand experience, and the technical knowledge to present challenges to TPTB's arguments when applicable, I can tell you its just too damn painful to do so.
No developer worth any salt is going to engage with him because it simply devolves into name calling if you don't align with his thoughts. Before long the n00b, b-lister, you are not worthy comments start regardless of if he's right or not and its just a noisy waste of time. The discussion ends in a blaze of fire with one of the parties leaving, usually it's not TPTB that has departed (he can argue all damn day) so it "looks" from the outside that he won, even if in reality his "facts" are wrong.
These days I'd rather go and pull my own teeth with a set of pliers than engage in any kind of discussion because its a less painful experience.
You could simply say where he is wrong, bringing infos, arguments and facts. I'm somebody without much knowledge about the tech, that's why I read discussions about it. My impression in general is (and over months) that there is a lack of controversial discussions about ETH. It's nearly always about the price. My questions about Casper, just for example, were never replied - in no topic.
This is the first thread I know of that focusses on the technical side and the lonterm-potential and if ETH can be what it is aimed to be. And I don't believe that one side is completely right and the other side is totally wrong, but what I see is: All the concerns in this thread are not rebutted yet.
He very well might be correct, but its a difficult task to figure it out as everything from him is a brain dump wall of text, with added shouting.
Not to mention that he changes his mind on things frequently, one day something isn't possible, then it is, then he has invented the best thing since slice bread relating to the issue at hand, then it isn't possible again.
Simply put, partitions for transactions are possible (we've done it, both with a block chain and with a ledger). Block chain based partitions are inefficient, ledger based partitions are not. ETH will be hard pressed to scale transactions with partitions on a block chain in excess of 2k+ tps, we've had public tests of partitioned ledgers nearing 3k tps (anyone is welcome to come and test it), and small WAN tests up to 10k+tps and I'm still not finished.
Partitions with scripts/contacts are a different kettle of fish, depending on the architecture, this may or may not be possible and the jury is still out on it.
The main issue is that you have a lot more data, or states, and have to ensure that there are no cross-partition conflicts of those states.
Further adding to the issue is that the states can be arbitrary, and a script/contract can have many of them. In the case of a transaction it is defined in the core protocol what the state of a transaction is and how that transaction can behave. With a script/contract you don't have that luxury as they can define their own states, sub-states and sub-protocols depending on the functionality.
Thank you for that answer and the further explanation! First: I don't care that much about personalities but sure, sometimes it can be a problem in discussions. In general I just try to filter out what's relevant.
I have to admit that I still don't understand it deeply but your explanation in the context of the discussion and my own researches about ETH seem to me like a confirmation that ETH is highly doubtable. I mean, it's a multi-million-dollar-project and there are obviously issues without solutions. Or with other words: It seems to me as if the ETH-team itself does not know yet how to solve these problems.
My own concerns were mainly about the eco-system if they switch to PoS. I know more about economy than about the technical side and I believe even if it would technically work, a PoS-System must lead into centralization and centralization would make the project irrelevant, impossible to trust it - even more when it comes to contracts. And if I'm right it has no chance in the real economy. Companies won't trust a system if it has no chance in longterm. Not totally sure if I'm right, maybe there are solutions to avoid centralization but if I understand this thread rudimentary correct, there are a lot more problems to solve.
And that in combination with the impression of a highly manipulated market... doesn't look good. As a speculator it's most likely possible to make some profit in the future, because of the hype, but if all those problems are real it's just a question of time that it's more widely discussed and more people would be careful with investing.
However, I'll continue to watch it and trying to understand if ETH has a chance or is doomed by design. But I won't invest any time soon.