Unfortunately I know as much as you about the tube in the article. In the case in Iceland the tube is fed through fissures in the lake bed, though there are also sinks in the bed of the river that drains the lake. In very wet conditions several of the caves around Laufbalavatn can become impassable. The tube could be any size, and it could also become choked with debris and silt or it could just lead to a lava seal and back up from there.
As I said in the other thread, this is not that uncommon, lava tubes quite often capture surface flow. The photo below shows a passage in Manjangul on Jeju Island (Republic of Korea), though the passage is dry in this shot, I am reliably informed that during Typhoons that the water level reaches the obvious line about 1m up the walls. Apparently the water is also shifting a bit too, not static. It would be quite a sight in those conditions.