Somewhat...
Ofcourse my post was at that time directed towards the constrained environment of the current patchlevel and not assuming one could design a whole new system from scratch.
Also the analogy with the banking algorithms arent quite appropriate since those mainly deal with trusted authenticated parties and with detecting any seeded data in a cryptographically secure link. This scenario does not quite apply to the gaming scenario where we either have a client spoofing data (more easily detectable) or a client presenting data in an unexpected fashion although the link itself is still secure.
The scenarios become increasingly problematic if the source is opensource, NOT because the algorithm itself is unsecure but mainly because the above mentioned presentation issues.
A really secure open-source environment would be based on
1) downloaded gamecode that includes random data, hashed on client and verified on server.
2) Data verification for video and audio based on a data sampling algorithm and verified on server.
Its important that all decisions are done on the serverside and that the data being used for the verification is unpredictable and non static.
As you can understand this is not something that is easily achieved and totally unrealistic with the current engines status but as machine grows more powerful and the CPU cycles are available it might be feasible. With the current tech it would be feasible using an old engine (such as ut/quake/quake2) but it puts some demands on the overall game engine design (presentation data that is constant no matter client settings related to sound and video).
I dont see this as anything happening soon or ever. Also I dont see the gaming industry caring about it since by the time the cheating becomes a HUGE problem the game is old and support is dropped anyway. On LAN or PRO event this is a smaller problem.
Anyway enough rant, public server cheating has always happened and will always happen, live with it and move on.
/e[N]eq