Can't speak for op, but can speak as just an observer - probably not. Blockchains are ultimately CP systems, not AP. From that link - "Data can be inserted into the sequence by appending the data to the previous generated state. The state, input data, and count are all published. Appending the input causes all future output to change unpredictably."
Meaning there is a single source of truth. That single source is decentralized (because blockchain), but it still requires a quorum to agree to accept a bit of data, and to treat the new state as the valid state. I.e., multiple competing parallel writes must be serialized across a network of computers, which is always more expensive than if they multiple competing parallel writes can be done in parallel.
Lamport clocks and etc are AP. It fully expects nodes to make decisions autonomously, and to synchronize after the fact. And data can be lost without realizing it.
(The above is all a bit of a simplification, but fundamentally, they're solving different problems)
Blockchains as commonly deployed seem to be more AP to me: you can get into split-brain and while there's a way to select the canonical form after the partition is healed, that's after the fact. Whereas e.g. Paxos won't even let you do anything without knowing it has quorum.
alas, because blockchains allow open participation, there is no reliable definition of split-brain, or network membership, or anything at all, really
because there is no canonical definition of the set of nodes which represent a source of truth for the state of the blockchain
the blockchain state is a function of game theoretical (read: unreliable and non-deterministic) algorithms that assume the result provided a majority (typically 2/3) of participating nodes can be assumed to be valid
they are I guess CP based on these assumptions (and others), but those assumptions are probabilistic, they absolutely can be violated, and those violations create forks of the chain which invalidate even the weakest consistency models
and nobody seems to care
the notion that blockchains are "trustless" is a complete fiction, all state necessarily exists in the context of some trusted authority, decentralization just means that this authority is not well defined, arbitrary from the perspective of its downstream consumers, unaccountable
Yeah, I should probably say "they're treated as CP systems", in that they purport CP guarantees, when in reality they adhere to almost none of them (but, they also adhere to none of the traditional expectations of an AP system, i.e., a way to heal partitions and restore some form of consistency).
Meaning there is a single source of truth. That single source is decentralized (because blockchain), but it still requires a quorum to agree to accept a bit of data, and to treat the new state as the valid state. I.e., multiple competing parallel writes must be serialized across a network of computers, which is always more expensive than if they multiple competing parallel writes can be done in parallel.
Lamport clocks and etc are AP. It fully expects nodes to make decisions autonomously, and to synchronize after the fact. And data can be lost without realizing it.
(The above is all a bit of a simplification, but fundamentally, they're solving different problems)