As the first initiative to any change of rules has to come from miners, we can say that the miner pool owners are somehow an oligarchy over bitcoin, but as they are part of a game-theoretical dynamics themselves, I don't know how to call this. It certainly is not a democracy. The "speculator/user" part has something of a free market to it. But only a few people can actually decide to do something, which turns it into an oligarchy, that is, a small set of aristocrats that can potentially change the rules ; however, with a market deciding over their acts.
Although I follow your reasoning and quite like the definition you gave, I have to argue this conclusion. In a censorship free market, you have no barrier to entry. You yourself can choose to be a Miner. This is much different than Oligarchy.
Yes, and no. On one hand, you are right, in that at first sight, there are no privileged aristocrats that can give the "right to mine", like they give licences to be a pharmacist or a wireless operator, and punish with violence those that try to do that without their agreement.
However, in reality, economies of scale and local subsidized electricity and manufacturing of asics make that there is a *de facto* oligarchy of mining pools, mining producing equipment, and profitable mining business. Essentially, as a small newcomer, you cannot enter, even though there is no aristocratic RULE with violence that holds you back. Call that "free market" if you wish. But if you push this logic to the extreme, then even states are a free market. After all, how does a state acquire its monopoly of violence over its territory ? By having an army. You can then argue that there is no barrier to entry: everybody can set up his own army, compete in the "revolution market", and try to win the de facto violence monopoly, so states are no aristocracies either if you push the logic you advance, to the extreme.
The point is that there is a kind of "winner takes all" system in bitcoin: the block that is won, makes all other competing work on the previous block moot. Yes, statistically, you can compete with small mining equipment, but in reality, you can't, because there's a collective unique consensus for each block, and there are no "small, local blocks with their own customers". In fact, that DOES exist, but it is called "alt coins". Within bitcoin, every block is a winner-takes-all competition. This is different in a real free decentralized market, where the local bakery can sell bread to the local customer, without a "winner takes all" global competition for each bread that is being sold.
The effective economies of scale that make mining only profitable if it is done in large pools, with large equipment, and with local subsidized electricity and cheap labor, makes that it does lead to a de facto oligarchy, even if that oligarchy is not a "privilege oligarchy" obtained by the direct licensing of the violence monopolist aristocracy.
The fact is, Bitcoin only rewards those who do things efficiently. Not wasting resources in inefficient tasks. This is exactly what a Free Market system does. It directs resources to the most efficient use. Everything else, would just be a subsidy paid by all others.
This is a bit ironic with bitcoin, because bitcoin mining is pure waste. You get in fact a competition of wasting where the "wasting function" is simply biased. In bitcoin, you have to abuse the bias in the wasting function to be "efficient" in the way you explain. Efficient usually means "More utility for the same waste", which is at the basis of the "best allocation of scarce resources". But in bitcoin, "proof of work" is simply "proof of waste", but the proof is biased. You can FAKE produce more "proof of waste" in certain circumstances than in others, but the "difficulty attained" is of no economic utility by itself. Nothing is produced. You only "proved waste", and some can "prove more waste with less waste" than others, which is the degenerate idea of "efficiency" in this particular case. If I can get cheaper electricity, and I can build smarter hardware (ASICS), I can fakely "prove" more wasted work than someone else, and I get to win the contest, but nothing more valuable was produced.
This is entirely different from "finding ways to produce cars with less resources". No, I find "ways to prove more wasted work with less resources", which begs the question. The proof of wasted work has no utility, contrary to cars that were produced.
The fact that you can get "economies of scale" and "efficiency advantages" in "proving wasted work" is simply an error in the proof of work function, which only provides a biased proxy to "wasted work", and doesn't prove any direct waste ; if it were, there couldn't be any efficiency differences, by assumption. Proof of work is (by definition) not useful. It was only meant as a tool to avoid sybil attacks, and to enhance true decentralization, while it degenerated in exactly the opposite because of the failed design of it.