Goal differential is an absolutely abysmal way to evaluate defensive play. As you yourself point out, different lines get different types of deployment, some more favorable than others. It's why we often see 3rd pairing Dmen on dominant teams have sky-high +/- numbers, despite being mediocre players at best.
Taylor Hall averaged 16:29 min of icetime in these playoffs, 8th among his own team's forwards. Largely because he was not as reliable defensively as most of the Canes' (admittedly excellent) forward corps. He wasn't used on the Canes' PK (which was outstanding, and one of the main reasons for their win) and wasn't used as often when the Canes needed to defend leads (which is what they did for most of the playoffs). Certainly the Canes' coaching staff didn't seem to feel like he was the most valuable player on the team. The thought of him winning the Smythe with this little icetime is a bit absurd; even Justin Williams, the worst Smythe pick of this millenium, played more per game than him. Stankoven I wouldn't have minded as much, because at least he's a more active, involved player, but I still think Staal was a more important player for this team.
It's the exact same case with Miller/Ghostisbehere. I actually like Ghost, but if you genuinely think he was more valuable to the Canes than Miller, all because he has 3 more points (which is little considering he was the team's PP1 guy), I question whether you're actually watching the games or just the stat sheet.
This is very reminiscent of the 2016 Conn Smythe discussion. And while I thought Letang would also have been deserving, I think they were right to give it to Crosby instead of Kessel, even if that 3rd line was terrific. The guy who handles the hardest assignments and wins them is more valuable to the team than the guys who feast with more favorable deployments.
