Peak level of play is not better. They all dominated their peers in a similar fashion, offensively, at their peaks.
Lower scoring levels by their peer groups doesn't change this.
I guess McDavid's 16/17 season was close to garbage too?
I’ll give your way of judging peak a go. We will look at number 2 in ppg and see how much they dominated. So with Crosby everyone points to the 99 games he played over the course of three seasons (11-13). We’d be comparing the points per game of 99 games to guys who close to or above 200 games for the span. That skews this wildly so we will look at each (very) shortened season at his peak level of play.
Crosby played 41 games from oct 7 2010 to Jan 5 2011 and averaged 1.61 to second place Stamkos at 1.37. Good for a ratio of 1.175. In 2011-12 he played 8 games from Nov 21 2011 to Dec 5 2011 and in that span he was in a four way tie at second in ppg with Ryan Kesler, Pavel Datsyuk and Ryan O’Reilly. He then played 14 games from Mar 15 2012 to Apr 7 2012 in which he was tied with Evgeni Malkin for first in ppg at 1.79. He was not the best player in the league this year and there isn’t even enough data to try to make the case (and over a peak Malkin at that). In 2013 he only played 36 of 48 games and overall he was first with 1.56 to second place St. Louis at 1.25 for what would be a 1.248 ratio but with 12 less games. In the actual span that Crosby played (Jan 19 to mar 30 2013) second place was Stamkos at 1.35 which is good for a 1.155 ratio similar to 2011. So for this span you have a combined 1.165 ratio over number 2 for 2011 and 2013 being the years you can count where he was clearly the best.
If you compare that to 2021 McDavid he is at 1.88 to second place (teammate) Draisaitl 1.50. A ratio of 1.253 over the full 56 game season. Next closest non teammate is Panarin at 1.38 which moves the ratio up to 1.362. 2023 McDavid has a ratio in the full 82 games 1.169 over second place Draisaitl again (1.87 to 1.60). Next closest non teammate was MacKinnon at 1.56 so the ratio would move (if you chose) to 1.198. A combined ratio of 1.211 for both years at 138 total games is clearly better than 1.165 over 77 combined games. Not only that but Stamkos was never the point producer Draisaitl or a MacKinnon are.
Jagr’s best year was 1999 in which he
dominated Selanne by a ratio of 1.097. Selanne also being better that year than Stamkos was at point production. McDavid’s peak is clearly above both by this metric and his 2021 is only marginally topped in the last 30 years by 1996 Lemieux at a 1.263 ratio. In a year like this Kucherov would be screwed by this logic and metric because in order to match the ratio over number 2 that Crosby had in 2011 he’d need to be averaging 2.06 points per game over a 1.75 McDavid. The quality of peers greatly matters.
As for McDavid’s 2017 it’s not garbage but obviously by no means close to his peak versions we have seen. He did come into a league that year that was low scoring as a 19-20 year old and put up 100 points. From the low scoring years of 2011 to 2017 in which league scoring never rose above 2.79 there were only 5 100 point seasons. Those being peak ‘12 Malkin, peak ‘16 Kane, prime ‘14 Crosby, peak ‘11 Daniel Sedin and a sophomore season ‘17 Connor McDavid who showed he could reach that mark as a player far from his own peak. Context matters there.