Many people point out that this type of tracking would only be possible if VPN companies handled logs to the Supreme Court, others say that patterns on the VPN connections could be detected and then the access to X be uncovered, at least partially (technically they only need to know you used X on a VPN, not what you saw on X or anything else). Others say that, besides logs, it is impossible to know if a citizen Y used X.
If you were, let's say, in the Supreme Court IT team and needed to detect who is using X through a VPN, how would you do it (besides asking the companies for the logs).
Subpeona the ISP for a list of users that have consistently connected to a range of known VPN provider IP ranges. Then they can filter by other heuristics like amount of data transferred and histograms that show when/how they used it. Unless the VPN provider is implementing advanced antitracking techniques or multihop gateways, it should be pretty easy to get that info since most ISPs are very much at the beck and call of the government.
Also, I saw many people saying that a read only access is fine, but posting could be more problematic.