Here's an alternative view of those seasons, split into 2 tables:
Rk | Name | Year | Team | Points | P% | VsX | | Avg VsX |
1 | Gordie Howe | 50-51 | DET | 86 | 0.364 | 130.30 | | 103.25 |
1 | Gordie Howe | 51-52 | DET | 86 | 0.400 | 124.64 | | 107.79 |
1 | Gordie Howe | 52-53 | DET | 95 | 0.428 | 155.74 | | 128.99 |
1 | Gordie Howe | 53-54 | DET | 81 | 0.424 | 132.79 | | 109.98 |
5 | Gordie Howe | 54-55 | DET | 62 | 0.304 | 83.78 | | 79.91 |
2 | Gordie Howe | 55-56 | DET | 79 | 0.432 | 111.27 | | 101.81 |
1 | Gordie Howe | 56-57 | DET | 89 | 0.449 | 115.58 | | 107.99 |
Rk | Name | Year | Team | Points | P% | VsX | | Avg VsX |
1 | Connor McDavid | 16-17 | EDM | 100 | 0.412 | 112.36 | | 102.29 |
1 | Connor McDavid | 17-18 | EDM | 108 | 0.472 | 105.88 | | 102.65 |
2 | Connor McDavid | 18-19 | EDM | 116 | 0.507 | 100.00 | | 108.45 |
2 | Connor McDavid | 19-20 | EDM | 97 | 0.435 | 100.00 | | 106.38 |
1 | Connor McDavid | 20-21 | EDM | 105 | 0.574 | 152.17 | | 145.17 |
1 | Connor McDavid | 21-22 | EDM | 123 | 0.432 | 106.96 | | 110.03 |
1 | Connor McDavid | 22-23 | EDM | 153 | 0.471 | 135.40 | | 135.28 |
It really comes down to how much weight to put on the shortened COVID year, because otherwise their peaks are absolutely similar - one outlier year way above average (52-53 128.99 for Howe, 22-23 135.28 for McDavid), and 5 years right around 6% above the VsX line, those seasons of Howe's adding up to 530.82, McDavid's to 529.8. Howe's 54-55 was disappointing relative to the rest of his peak, but he did miss time that year. On the other hand, the stats back up that McDavid was running hot in terms of the random aspects of points (oiSh% and IPP) in that COVID year, and the extra 26 games to a full 82 would have featured some regression. [He got points on 68 of the 74 even strength goals he was on-ice for, when it probably should've been 54 of 68ish.]
When looking at scoring in NHL history, it is remarkable that P% does not change - an elite season is 45%, plus or minus 5% (more like 42.5% plus or minus 7.5%, if you want to get more precise), whether it happens in 1953 or 2023, and any year in between. It's the same thing with goal scorers - a peak year is 20%, plus or minus 5%, and it's indistinguishable between Maurice Richard and Alex Ovechkin. The only thing scoring levels affect are the raw numbers, and even that is masked a bit by games played. League average was 163 in 48-49, 190 in 50-51 and 223 in 16-17, but per-82 games, all those years are 223. So Howe's 50-51 season of 86 points equates to 100.94 points in 16-17, the difference being almost entirely the extra 12 games played, with a small fudge for rounding. You can set league average to be almost any arbitrary number that makes sense in terms of games played/goals per game, and the %LA/G%/P% will map on to it, and return numbers that make perfect sense (apart from goal/assist balance issues for pre-consolidation players).
I do think there are a few factors pushing point totals up now compared to Howe's era. First off, the records Howe was pursuing were set in 20-25 fewer games, so there weren't any real targets for Howe to set as goals. Seventy years later, McDavid has all the targets he could want. Secondly, 6 teams in a talent-deficient era meant that Howe had little competition in his peak years in the point race, so he had little incentive to try and boost his totals. With 32 teams, there's a whole lot more player-seasons to have outliers. In those 7 Howe seasons, only Beliveau in 55-56 and Lindsay and Beliveau in 56-57 exceeded the 100 point VsX threshold (apart from the 6 Howe years). In McDavid's seasons, his 7 exceeded 100, along with 10 other seasons (Kucherov 18-19, Kane 18-19, Draisaitl 19-20 20-21 22-23, Panarin 19-20, Pastrnak 19-20, MacKinnon 19-20, Huberdeau 21-22, Gaudreau 21-22 [not to mention another 5 seasons in the 98-100 range that just missed]).
In the end, I'd say that Howe and McDavid peaked at the same level, but Howe has 14 more years of elite production that McDavid hasn't had the time to match.