deck penetration in KO system

[Heraclitus]

Just tried Vancura and Fuch's "KO system" last night for the first time with good success (well, maybe mostly luck, but at least I won. But a question occurs to me about the authors' deck penetration assumption. The authors in Appendix 5 say they assume 75% deck penetration and that although lesser penetration is unfavorable to the player it does not affect how to use the system itself.

However, I'm interested in what happens as one approaches the end of the shoe. Assume 8 decks. With the prescribed -28 IRC and -6 key count, the odds have already become slightly unfavorable at 75%: The average count at 75% is -28 + (.75 x 32) or -4. Thus (-6 + 4)/2 = -1. To be neutral at 75% penetration entails a -4 key count. Perhaps the difference is negligible--and I realize that the idea with the KO system is to not have to count decks remaining. But I'm wondering if it would make sense to gradually reduce the key count from -6 to -4 over the course of a shoe. Or is the difference too negligible to worry about? I was surprised last night how deep the deck penetration was (at Foxwoods, CT). Seemed over 80% to me, although I'm not practiced at the estimation. But for one and a half decks remaining (81.25%) the neutral count is -2. That difference seems significant to me.