I looked at this briefly last week, but I wanted to offer slightly more data. Here are team averages over the last four seasons in the last five minutes of regulation:
|For, goalie pulled||18.68||11.8||100.9||11.6|
|Vs, goalie pulled||18.68||22.0||40.1||54.8|
I think the shooting percentage on the empty net is garbage - somehow a number of shots on goal that didn't go in were recorded. And how do you differentiate between missed shots and icing? But the shots for are good, as are the goal-scoring rates. I've gone into this before, but pulling the goalie massively increases a team's likelihood of scoring (to roughly 20% in each minute of empty net play), and increases the likelihood of getting scored on even more.
I think the numbers for "in-game" empty net time - that is, when a team pulls its goalie for a delayed penalty call - are more interesting:
|For, Goalie Pulled||5.2||10.5||147.2||7.1|
The shot rate in these situations is 50% higher than it is during end-of-game empty net situations, but I don't think this reflects additional defensive passivity while waiting for a penalty to be called. It's just a consequence of play ending as soon as the defensive team touches the puck; at the end of the game, the defending team can clear the puck out of the zone, shoot it down the ice or score - all activities that eat up clock time and prevent the offensive team from getting a shot on goal.
Incidentally, do note how much ice time each team spends with an empty net: 14 seconds per game at the end of the game; 4 seconds during the game. How an individual team does with an empty net reflects very little on their talent and has no predictive value for future empty net situations.