Layers, not a single question
One of the biggest mistakes in reasoning about blockchain law is treating it as one question with one answer, 'is this legal?', instead of a stack of separate, layered questions that each need their own analysis. Is the smart contract's execution likely to be treated as reflecting a real legal agreement between identifiable parties? If something goes wrong, a bug, a hack, an unexpected outcome, is there a real chance a court or the community intervenes rather than treating the code's output as final? If the project operates as a DAO, does it have a legal wrapper, and if not, are participants exposed to something like general partnership liability? Which countries can plausibly assert jurisdiction, given where contributors live, where users are, and which financial rails move funds in and out? Does the token's design and marketing risk it being treated as a security? And separately from all of that, do specific people, frontend operators, active contributors, still carry personal exposure regardless of how decentralized the system claims to be?
None of these questions has a universal answer that applies to every project, and that's exactly the point: 'is my project legal' isn't answerable in the abstract, the honest answer is almost always 'it depends on jurisdiction, structure, and specific facts', which is a genuinely different and more useful thing to understand than a false sense of certainty in either direction. Builders who treat every one of these layers as automatically fine because 'it's decentralized' are working from a mental model that doesn't match how legal systems have actually behaved. Builders who treat every layer as automatically doomed are usually overcorrecting into paralysis that isn't justified either.