I'm not sure the parent comment is necessarily going for a patchless solution, and as you say, by the time you have got far enough with reverse engineering for the particular type of encryption to matter, you can always just patch the executable to bypass it altogether. (In fact by halfway down the article the author had done exactly that.)
Besides, I'm not even convinced that "any competent company" would bother with public/private key cryptography given that it makes little difference to them how exactly their copy protection is broken.
The bigger problem for the parent commenter is if actual game logic is executing on the server, which is probably the case for online multiplayer games.