I think this is one of those things where there is no right or wrong choice. its just a business decision. people will vote their support with their wallets, as players and investors. Only time will tell if this change is significant in changing things (business wise), or if it doesnt change anything at all.
i'm not too sure either way.
The main difference is that the new way increases variance for investors, and reduces it (to zero) for the house.
When the site loses, investors lose to the site and to the winning players.
When the site wins more than expected, the players win from the losing players and pay less than 10% commission on their profits.
So the effect of the change (for investors) is that losses are worse and wins are better than before.
The expected result is the same (assuming the different edges for the various games have been taken into account correctly; someone mentioned that blackjack's edge isn't 1%, and so taking 0.1% of turnover isn't right for blackjack bets).
My argument was the fact that 90% of the sites are much below the profit of 1% wagered and not every investor plans to stay invested in the long run. So people who have to invest are actually risking more that the site remains in profit of 1%, after they invest.
For example if people invested in dice.ninja, and this investment system was implemented then the investors would have paid around 20-25% extra of their losses.
And same is the case with dice now , the total amount wagered is around 22.8K , and expected is around 228 BTC, but they are currently at 30 BTC profit, out of which commission to the site would have been 22.8 BTC so basically profits would just be around 7 BTC, and they will be losing much more.