1MB4EVA is not being proposed as a serious technical solution. On h/ware improvements alone capacity will grow but how to implement that in a secure algorithm is the challenge, given the myriad of parameters and delicate balance of the existing competing incentive structure.
It would not surprise me that after all the work that is going on now for secure transactions anchored in main chain, but off-chain, that it turns out blocks wont be demanded by the market to be much bigger at all. Highly distributed networks scale securely best self-similarly, not monolithically (simply bigger blocks is a monolithic approach, so likely wrong) and "big" is a temporally-relative term that is near to impossible to encode in network rules that need to persist for decades.
I don't have to be an expert to recognize technobabble when I see it. You're like a crooked mechanic telling me I need my muffler bearings changed and my headlight fluid topped off. It's too late for fucking blocktrees or whatever. We're getting full blocks NOW. You think 8MB is drastic but you want to do something even more radical? Like replacing the blockchain with trees or multiple chains? Is that your idea of being careful?
It must take a genius to be that stupid.