Got to be careful with the crayola analytics without proper context, which happens way too often.
Of course the regular season offense was going to be better this year than last. We added more offensive talent (Bennett full season, Reinhart, Lundell, Giroux at the end, March upgraded) and ran with a wide open system in a season when offense was lights out across the league. I know the numbers are relative, but the play style across the league definitely favoured our team make up.
The powerplay metrics improved in the Tampa series after a ridiculously long streak of futility that even Bruno admits he was too stubborn about changing things up. So yes of course they look better, because they included the changes that everyone was saying should have been done earlier, including himself in hindsight. Are we really giving credit for changing up something that should have been changed way earlier before giving up a 2-0 lead? And yeah of course Vasi was better than the scrubs in Washington, that's kind of beside the point and not addressing the issue.
Also, looking at unblocked shot attempts is a dumb way to look at power play effectiveness. It makes a lot of sense for 5v5 play to take out the noise and randomness with blocked shots, plus the other team blocking a shot means they don't have possession. But it's dumb to look at unblocked shots for the power play because not getting your shot blocked (or blocking a shot for the pk team) is a hugely important aspect of being effective on the power play. We were getting shots on the PP blocked like crazy, that was part of the problem which I don't think shows up in that chart.