Marcel Dionne was in his prime when the NHL was at its highest point ever for scoring. He was sixth among centers in PPG and 11th overall with 1.31 during his career. Among the notable non-HHOFers: Kent Nilsson (17th, 1.25), longtime teammate Bernie Nicholls (aka "the best second-line center ever", 19th, 1.23), short-term teammate Jimmy Carson (aka "the reason LA traded Marcel Dionne", 22nd, 1.19), linemate Dave Taylor (32nd, 1.12), linemate Charlie Simmer (54th, 1.00), Brian Propp (35th, 1.09), Real Cloutier (38th, 1.09).
It's amazing how many of those guys are offense-only players who also played with Dionne. Almost like he might have benefited from playing in an offensive system full of skilled offense-first guys.
You used that argument on Fedorov, who played in a defensive system and in a defensive role. Yet you don't see that it's actually applicable to Dionne? I'm not saying Fedorov would have scored more than Dionne in the same situation (who knows, maybe he scores 140-160, maybe he stays around 100). But what I am saying is that Dionne had an environment much closer to what Fedorov had in 1993-94 than 1998-99. Closer in linemate production, closer in overall league average scoring, closer in league goaltender quality. Dionne still played in a higher scoring era, and had higher scoring linemates; Ciccarelli wasn't as high-scoring as either Simmer or Taylor during the given period, and he was basically washed up by the time Fedorov had him. Kozlov only three times in his career broke 0.90 PPG (only once with Detroit), and never hit PPG (he did come one point away with Atlanta centering Hossa and Kovalchuk on the PP, though).
Dionne vs. Selanne would be an interesting comparison. I would probably lean Dionne, but it would make for an interesting argument.