Well,
I think you understood what I was trying to say very well.
Lets say a customer loses 100 btc, you charge investors 0.5 btc. Now the customer wins 100 btc back. Result: customer is breakeven, investors lost 0.5 btc, you gained 0.5 btc.
Yolodice had a similar system with with 15% and then 25% fee IIRC, but only took the fee once a week, which makes it much better for investors.
We're on the same page with regards to what you meant, but my point didn't come across well: Your example could happen with the old commission on wagers as well. If a player loses 100 BTC and wins it again, the player would have broken even and investors would have paid at least 0.25 BTC in commission to bustabit. More so, unlike with a commission on profits it was possible for investors to have paid a commission without ever having earned a net profit at all. For example, if a player won 100 BTC without losing it again, investors would have been down 100 BTC and still paid the commission.
So while increasing the commission isn't, switching to a commission on profits vs wagers itself is strictly better for investors.