You can easily prevent wallhacks; just don't tell clients information they shouldn't display to the user, such as the positions of players the user can't currently see.
Also you may be able to hear the person behind the wall , for (even though it may only be a very quiet sound from the other character walking slowly). The client application will need a location in order to play the sound correctly, this information could be used to make a wallhack.
The other alternative would be to pre-prepare all sound on the server and stream it to the client , but this would put extra stress on the server and more demand on bandwidth.
If the data was received encrypted, you could have a relatively small decryption key sent when necessary, thus solving the "prefetch from server" problem.
If the model files were broken up into chunks and xored with each other in various ways, and saved in a redundant manner, I could imagine that there is a setup in which there are a combinatorially large number of possible ways to fetch a set of blocks which reconstruct a given model, and if you prefetch a few red-herrings as well, it would be infeasible to determine which model you are loading based just on the commands the sever sent you on which blocks to load. Then, when the time comes, they send a very short command which tells you which blocks to use and how to reconstruct the model. Hmm... but now I'm thinking about it, I think you'd need a large overhead in both directions to make it difficult to just match the patterns of which blocks you used.
Online games always walk a tightrope between performance and security. The currently popular model seems to be to optimise the game for performance but use a second program to verify that the client code has not been modified.
Of course you then have to make sure the anti-cheat is not modified itself. I have to say though that cheating is not a big apparent problem in most online games that I have played. Most cheating problems are in older games that have not been updated in some time.
It's easier to trust a "cheating hack" if it comes as a source code patch or the source code is readily available. If I was so inclined to cheat, I certainly would not trust a strange patch/executable from some shadowy figure I knew nothing about.
Technology will eventually solve such problems (thin clients or fully homomorphic encryption) but these have serious issues of their own.
Not at all, as it is anybody can modify the game binaries in the same way they would modify the source code (albeit with more difficulty).
Cheating is prevented by having a central server validate all client actions for consistency, or in games without a simulation server (Starcraft II) having clients compare a game state hash and disconnect if they differ.
Client side cheating (Wallhacks, Aimbots) are not preventable without a locked down client, open source or not.