X, A, B and C all trade independently and are subject to their own buy and sell pressures that affects their prices. But the price of the ETF tends to be consistent with the price of the underlying shares, for a couple of reasons.
The fundamental value of X is just the aggregated value of the underlying shares. After all it's just (indirect) ownership of shares of A B and C. Who would buy X if it's trading at a large premium compared to A, B and C?
But, more importantly, authorized participants can hand in one share of X to the ETF provider and receives shares of A, B and C in exchange. They can do the opposite too, they can hand in shares of A, B and C to receive one new share of X.
Therefore, if X trades at a discount compared to A, B and C, an authorized market maker can make a profit by buying shares of X, redeeming them for A, B and C, and selling those. Which will drive the price of X up, and the prices of A, B and C down, thus causing them to converge. In this way the sell pressure on X is "passed through" to A, B and C. The reverse happens if X trades at a premium.
This process is not instantaneous or friction-less so the price of the ETF will not necessarily match the underlying shares exactly at any given time. But significant differences tend to be arbitraged away.
I would say it's both true that the price of X affects the prices of A, B and C, and the prices of A, B and C affect the price of X. The prices will tend to converge because of arbitrage but in general it's difficult to say which caused which.
Of course the relationship is not quite symmetrical, because X derives it's fundamental value from A, B and C, the reverse is not true.