A lot of these results depend on conditions, and the results are often close enough that the conditions can change which one is "better." For my standard game I use a S17, 8D shoe with 1.5 deck pen, and Wonging out. This I choose as my standard because it's inevitably the kind of game an East Coaster ends up playing. So it's possible that both my numbers and Cacarulo's are correct given the different parameters. But I have just never heard of Hi-Opt II without the sidecount outperforming anything; that is usually considered an unwise way to play.
For my kind of game HO2+A is a couple of percent better than Mentor, which is a fraction of a percent better than RPC (which I use), which is a percent better than Halves. Zen and High-Low somewhat underperform these. You will spend several lifetimes at the table waiting to see the difference between the first four.