We had an extension (web and mobile) called Sterling with monthly subscription tiers. It would automatically block ads on all sites and signin for paywalled sites that joined the network, or we would use some sneaky techniques to gain access to the article anyway (api/reader views, etc) to at least test the viability.

Each users money was split by time spent on each domain monthly which was the only collected data. Held in escrow until publishers claimed each domain. Biggest problem was getting people to consider that subscriptions were worth it, even if they were spending 10s of hours on these sites. $0.99 had interest but was not viable. $9.99 had some uptake but under 1% conversions.

Overall didn't make enough money and pubs also too scared/proud to work on anything other than their own subscriptions. Lots of politics in this industry and lack of cooperation is a major obstacle.