Went back and looked at Barry Bonds' '01-'04 years and Ruth's '20-'21. It's a good comparison set to see the merits and flaws of common baseball statistics.
In '04, at the age of 39, Bonds posted a .610 OBP and .812 SLG for an OPS of 1.422. That .610 OBP is the highest ever, and the OPS of 1.422 is the highest ever. In '01, at the age of 36, Bonds posted a .531 OBP and .863 SLG for an OPS of 1.395. That .863 SLG is the highest ever recorded. So, according to baseball metric types, Bonds' '04 > '01 by a bit, because OPS.
Except if you look at bases per out, Bonds in '01 posted a 1.40 BPO vs. 1.31 in '04. Basically, he made 61 more outs but posted >100 more bases. (There are some calculation differences depending on how you handle HBP and SF, but they don't change the relative math.) The 1.4 BPO in '01 is the highest ever.
Contrast Ruth's two best statistical seasons of '20 and '21.
In '20, Ruth posted a .527 OBP, .847 SLG and 1.37 OPS. That SLG is the second highest ever recorded next to Bonds' '01, IIRC. In '21, Ruth posted a .510 OBP and .846 SLG, for an OPS of 1.356. In '21, Ruth posted 457 total bases, incidentally, which is the highest total bases season in baseball history.
According to the baseball 'stat' types, OPS is the metric and Ruth's '20 season was better than his '21, and below both Bonds' '01 and '04 seasons. But if you look at bases per out, Ruth posted 1.36 in '20 and 1.37 in '21. By that metric, Ruth's '21 was better than his '20, and his '20 and '21 were better than Bonds' '04 but not quite as good as Bonds' '01. (Same adjustment methods for both guys, each season.)
I'm pretty sure basic math will tell you that if you run the metrics through a simulator, BPO optimization wins over time, every time. I believe in any case, those are the top 4 offensive seasons ever in baseball, measured by OPS or BPO.