If Gronbog can simulate T count, it's safe to say it's little he can't do. There was an immense amount of work on that project. Verifying the values being plugged was as difficult and detail-oriented as it gets, considerably more trouble than actually learning or performing the method, which is from a practical application standpoint the maximum gain and efficiency with the least amount of mental strain, arriving at near perfect play. That's what it takes to outperform HiOpt2ASC and anything short of that isn't going to cut it. I'm not saying what you're doing isn't good or effective, or discourage you in any way, and merely saying it's relatively certain it won't outperform HiOptASC, with this coming from someone who has intimate familiarity with what it takes to outperform HiOpt2ASC. Gronbog doing a simulation for you would give you a definitive answer on your enhancements compared to everything else, much better than just speculation.

Most of your money made from card counting is derived from your bet spread and optimal betting. Everything after that is just gravy, slight gains that are tiny fractions of a percentage point, worth going for with some, not worth bothering with for others. The illustrious 18 was devised for maximum gain using the least effort, getting the most of it. The tiniest fraction of a percentage point is a lot if there's a large enough volume of money involved, though, so nothing wrong with trying to maximize your theoretical edge. You can run this out to its greatest extent even, with the most difficult composition dependent index possible for the least gain involved if you want to, where the difference in EV can be as little as .0001 between three possible playing decisions, to either hit, split, or double given the right deck composition, which will have virtually no impact on a simulation, but is still just a hell of a lot of fun if you are hardcore enough. Can you guess the hand?