China is the world's largest crude importer and its second largest consumer. That alone would make it central to the market. What makes it analytically difficult is the combination: it is the most important single source of demand-side variation and the least observable one.
The structure is unlike a purely commercial market in three ways. Refining is split between large state-owned enterprises and a substantial independent sector, the teapot refineries, which respond to margin far more sharply than the majors do. Crude imports are partly commercial and partly strategic, with reserve buying that is neither published nor predictable. And product exports are governed by an administrative quota system rather than by arbitrage alone.
Each of those introduces a wedge between price and behaviour. A commercial refiner buys when margins justify it; a state buyer accumulating reserves buys when policy says so, which can mean buying into weakness. A refiner in an open market exports when the arbitrage opens; a Chinese refiner exports when it has quota. So the usual inference from price to flow is weaker here than anywhere else of comparable size.
The practical consequence for a desk is that Chinese demand is usually inferred rather than observed, via an implied-demand calculation: refinery runs plus net product imports minus product stock change. Every term in that has its own error, and the result is a derived number with a wide band. It is still the best available read, and it should carry its uncertainty visibly rather than be quoted as a measurement.