During defi summer I generally posited a thesis of BTC + defi apps under the general idea that you can abstract out ETH from defi apps (aave, just use wbtc) and that BTC would crowd out ETH on its own chain Later changed my mind when I saw the demand for protocol-native SoV The rise of CeDeFi has been the most bearish thing for ETH actually bc it means that a huge % of business logic that ppl theorised would become open smart contracts running on distributed networks, actually are closed source code running on single servers Ethereum does need to embrace use cases that benefit from global distributed compute and censorship resistance. Some of these are indeed more cypherpunk than a lot of its current supporters are comfortable with--things like privacy for instance. Others are actually just lower hanging fruit that ppl havent gotten around to building I dont think you can really have a DeFi resurgence if ETH itself doesnt do well though, bc in a world where there's low demand for validator decentralisation and smart contracts, you can just do stuff offchain and save time and money