Standards are boring, and that is the point
The best token standards do one narrow thing well. Every attempt to make a 'complete' standard has collapsed under its own weight.
Token standards that try to solve the whole problem — token interface, asset model, compliance logic, waterfall semantics, everything — have a remarkable track record of collapsing under their own weight. Token standards that try to solve one narrow problem very well — just the token interface, or just a compliance hook — have a remarkable track record of surviving and compounding.
ERC-20 did one thing: transferable fungible tokens. It did not try to specify what the tokens represent, how they are issued, what legal obligations attach, or what compliance logic applies. A decade later, it is the substrate on which nearly all subsequent standards were built.
ERC-3643 did one thing: a compliance-hook interface that lets an issuer plug in per-holder eligibility checks. It did not try to specify the eligibility logic itself, the KYC process behind it, or the legal framework in which it operates. It is the most widely-adopted compliance-aware standard because it is the least opinionated.
ERC-4626 did one thing: a tokenised vault interface. It is enormously useful because it refuses to specify what the vault contains or how the vault's yield is generated.
The pattern is consistent: narrow standards that compose well beat comprehensive standards that don't. This is not a tokenisation-specific observation. It is the IETF model of standards work, and it works in financial infrastructure for the same reason it works in networking.
When the Association's Asset Standards working group considers its output, we are applying this lesson deliberately. The draft we are circulating is narrow — just the asset-model schema, not the token standard or the compliance logic — and we are resisting the pressure to expand its scope. The pressure is real and it will keep coming. The discipline is worth preserving.