Power Play - Shots vs. "Shot Quality"

This is one of the stat side of things that I either don’t understand or don’t agree with. I can see the relationship between shots generated and powerplay goals. But obviously the quality of the shot plays a role in how many goals actually go in. Is the arguement that over time high quality chances are unsustainable?
--Peacecountry, in the comments .

A quick check of the data from the last four years via the venerable and terrifying Gabriel Desjardins shows the relationship between shots per sixty minutes of 5v4 time on ice and power play success (power play conversion percentage) is extremely strong, with an r^2 = .400 So, given our sample size, 5v4 shots per 60 predicts 40% of power play success.

But the Oilers are 24th in the league in shots per 60, which is a mountain to climb if they want to have a top 10 power play. Put simply, four teams in the last four years have had a top ten power play by percentage and finished in the bottom ten of the league in shots for per 60 with the man advantage. The teams to pull off this feat were Philadelphia and St. Louis in 2008-09, Minnesota in 2009-10 and St. Louis in 2010-11.

St. Louis ranked 29th in SF/60 in 2007-08, 24th in 2008-09, 15th in 2009-10 and 23rd in 2010-11. Mr. Desjardins has shown that a team's power play shooting percentage in half of its games has almost no predictive ability for guessing how it will do in the other half, so we should attribute a high shooting percentage to luck rather than a scheme that produces high quality shots.

If the argument for shot quality on the power play rests on St. Louis, we must assume they forgot how to take quality shots in two out of these four seasons.

Another way of looking at this: over the last four years, there have been 120 individual seasons. Of the top half of all power plays during that time (60 of 120) only eight of those teams finished in the bottom ten in shots for per 60 with the man advantage. While there is gray area in there (some top 10 teams obviously finished in the middle 10 in shots), the jump from the bottom in shots to the top in conversion is a difficult one.

It's possible to have a top ten power play with bottom ten shots generated, but it's not desirable, nor is it necessarily a product of shot quality, but more likely a product of variance. So, yes, Peacecountry, the Oilers might finish in the top 10, but it's going to be based on luck, not shot-quality.