Yes, I see that occasionally. In that case, I would think that the first thing to freeze and break would be the nozzle, which would be fully exposed, rather than the sillcock tube within the building envelope. That still leaves my question unanswered: "Why would a hose still attached to a frost free sill cock cause a pipe to burst?" True, the nozzle will probably be the first thing to freeze, but it won't leak until it thaws. So the freezing continues back up the hose, sill-cock tube, and pipe until temperatures warm. Then when it thaws, all of the little splits that the expanded ice made, begin to leak and spray until the water pressure is depleted. The splits closest to the water pressure source will be the first to leak the most. The nozzle is at the far end of all the frozen ice and won't leak to its full potential until all the ice in the hose has been melted. This can take a while as the ice will push down stream as it melts plugging off or slowing the leak each time the ice shifts down stream. By the time the hose is fully clear, plenty of water has escaped the first split in the system.