Not very good for shoes
In 6-deck S17 DAS DOA, you can squeeze out slightly better than even if you use a large enough spread (about 12:1 or more), but the problem is a very large variance for a very tiny edge. (Of course, it's still better than playing a progression ;-)
I got interested in unbalanced systems from reading KO, too, and then from reading some of Brett Harris' posts on this board. Lately, I've been using KO in shoe games with the IRCs and keys adjusted downward by 4 and the pivot at 0. Most of the time, I just use the running count and the key, but for close situations like reaching the key early or late in the shoe, it's very easy to do a true count conversion to check it. I really think that Fuchs and Vancura should have defined KO this way, as an easy transition into TKO, beyond KO Prefered. Really, the only reason they defined it with the pivot at +4 was to make the pivot advantage close to that for Hi-Lo +4.
The pivot for an unbalanced system is just the net imbalance per deck times the number of decks. That (by definition) will be the only count that has a known advantage, at any point that it's reached in the deck. With this A-4-5 system, the +4 pivot will be somewhere around +1.5% advantage and up, depending on the starting disadvantage (which makes it pretty close to Hi-Lo +4).
The key count is determined by sims, and it really represents an average over the "typical" game rules and penetration. I got the numbers I posted using CVSim, which has a report showing the advantage at each running count number. The key is just the lowest running count that shows a definite advantage. (The KO keys seem to have about 0.25% advantage, on average of course.)