I have ran simulation for this cases, and don't see any issue with it.... Am I missing something....
you claim to have run a simulation with an adequate sample size to prove each of these cases within minutes? an adequate sample size to prove these would be hundreds of thousands of bets, and I doubt your small server is capable of handling that. yes, you are missing something, its called math.
I ran only first 3 payouts on my own simulator with 1 million bets made. Result was the same.
But that one is interesting with payout rounding, yeah. Thanks for pointing that one out.