The goal is to build a permissionless and censorship-resistent financial network. This has been going extremely well. People referencing their engineering credentials and arguing that the "tech is garbage" always have lots of opinions on the usefulness of such a permissionless and censorship-resistent financial network (not a tech question), but are notably silent as to which technical decisions where incorrect, and what the correct ones are. The latter is key, because unless there is a better way to build for the desired goal, there is no tech-deficiency here.
And to put the timeline into context: Note that AMMs like Uniswap, a key infrastructure innovation in the space, are about 4 years old - before this, it was basically impossible to exchange two tokens in a decentralized way. Other key pieces like cross-chain communication systems and rollups are still being built.