Am I right in thinking with this counting system, one must keep the running count of the deck, and three separate side counts for three other cards?
Seems an awful lot of strain on the brain, and I would think this system would need to have very siginificant advantages over hi-lo to justify the extra work and higher risks that go with getting one of the counts wrong - ie overbetting.
Playing Hi-lo is all I need at my gaming level, which is typically the equivalent of $5 or $10. Now if I were a pro/semi-pro, betting £50/£100 per hand, I might be inclined to look at something more efficient (but that's never going to happen). I have considered to returning to use an unbalanced count, but why bother? For me, it'd be the equivalent of keep trading in a car for something marginally more economical when I only drive a few hundred miles a year.
Because of this I've no real interest in spending time making a detailed analysis of this gentleman's system. The fact that it has side-counts means it probably not a good one for the average recreational player who wants to counter the house's advantage, unless of course they have masochistic tendencies.
Each to one's own. If it doesn't look like it's up to much, don't use it. If you're new to all this counting stuff, use something simpler (to begin with at least). I'm a great believer in the adage "if it ain't broke, don't fix it".
On the question of computer simulation, I'm sure it wouldn't be too difficult to create a program that allocates a unique identity to each of the 312 cards in a shoe, random shuffles and then varies the bet amount dependent on the remaining composition (312 less specific cards dealt and cutting off the last 78 cards). This would settle the debate once and for all. In the absence of a ramp, quarter Kelly betting could be assumed, although with the side counts it isn't easily obvious how big the advantage actually is when it turns up.
Just a few thoughts.