I think if we purposefully try to crash / break the bitcoin testnet network via as much attacks as we can, and see how the network reacts, we could learn a lot of valuable information on how to secure the bitcoin network against attacks.
Seems like doing a bunch of stuff at once would not be the way to learn what broke it. If you have any ideas give it a try.