[1] https://upswingpoker.com/ultimate-bet-absolute-poker-scandal...
> The cheaters relied on a superuser account named “Auditmonster2,” that would observe tables and was able to see everyone player’s hole cards.
So.. the game was played exactly by the rules, and the full visibility would not have helped. After all, it was purely info exposure.
But you may say: in blockchain, we'd design special cryptosystem that will ensure other players do know know cards too early, as well as a verify that game is played according to the rules
My answer would be that if you have designed such a cryptosystem, you don't need blockchain anymore! Keep existing infrastructure, but apply encryption and verification at each client. After all, you do not care about world consensus for your poker game -- you just care about consensus between players. You will get all the advantages of centralized solution (high speed, very scaleable, low device requirements, easy upgrades), and server won't be able to rig the games.
Blockchains are mutation-resistant, but not immutable. For example, here's the cost of performing one-hour "51% attack" on various PoW-based blockchains: https://www.crypto51.app/
Because of the relatively small audience that would dedicate their computers to running this vendor's blockchain 24/7, the cost of brute-force attacking this game's blockchain would likely be much lower than any of those. That assumes the software running on each node was unhackable, which would potentially make it much easier to alter or destroy.
> Cheating would be very hard…
At best, a distributed solution doesn't make cheating more difficult. Depending on how it's implemented, a distributed solution can make cheating and other abuse much easier (the idea of community managers banning bad actors isn't a thing, etc.).
I don't know about being 'freed from the constraints of manufacturing'.. But a game on a blockchain might be one of the few uses of blockchain I can think of that actually sounds useful.