Nick, what I suggest you do is write some blackjack simulation software.
This software would need to do the following things:
Create a shuffled deck of X decks (X should be configurable at runtime)
Deal out a dealer hand and Y player hands (Y is configurable at runtime)
Play those hands out according to blackjack rules, calculating and recording the result of each possible player decision.
Then you run that simulation a few billion times. Run it 1 billion times with 1 player hand, then another billion with 2 player hands, etc.
At the end, you will be provided with:
the correct decision to make in every situation (strategy charts)
the overall HA
You will then be able to test if there is any difference in HA if you have more or less player hands, and exactly which decisions give you the best return in the long run.
If you are incapable of writing this software, I suggest you stop arguing and accept the results generated by those who do understand how to simulate this. People like the Wizarde of Odds, Stanford Wong, etc, who all say that:
1. HA is unaffected by the number of hands you play
2. You should never ever split tens.
(PS. Your assertation that playing more hands affects return may be based on card counters. It is true that professional blackjack card counters often move from playing 1 hand to 2 or more. That is because they wait until the current deck state is providing them with an edge, and then they increase their bet size to take advantage of that edge. This is called the ‘spread’. One way of spreading is just to increase the bet on one hand. The other is to add more hands.)