My answer is known here, but I'll repeat for some of the newcommers - UNLESS you use proficiently a secondary count consisting of A-2 vs 2&5+1, or similar, you will not realize the extra theoretical gain that AO2 portends. Ace density estimate per 1/4D as proposed by Humble and Carlson and Uston and Revere will NOT cut it. SO you wind up working harder to get the same gain as ZEN w/o Ace side-count. AND if your Ace 1/4 deck density estimate is not up to speed you won't even get your ZEN's worth.
If you disagree with me on this you are disagreeing with Arnold Snyder, and especially with Ken Uston, who had a serious vested interest in Ace side-count systems.
Now IF you only play 1-2D and you must side count a single card, use a modified ZEN (ie, 12222000-2-1) with a side count of 7s with bivaluate departure matrices.
Oh, and by obsolete, perhaps I should say instead 'antiquated'.
Lastly, AO2 is further antiquated by virtue of the inclusion of the 9-1, which is overall detrimental and offsets at least part of the theoretical gain from the neutralized Ace. And which is why HO2 is stronger than AO2. So there is a two part improvement by swapping the A-9 tags in AO2. The good news is that you can keep your AO2 indices. zg