If you want to learn about impulses it will be helpful to learn about the latest dynamic convolution technology.
Speakers are reactive. Single impulses don't address that and that is why most popular speaker sims sound stale.
If you send some lively, uncompressed guitar slinging into a static impulse it will be obvious how it becomes conditioned and un-lively.
It is less obvious that the sound becomes stale if you send it a less lively signal.
The latest implementations of dynamic convolution processors use multiple impulses of specific speakers made at varying intensities and states in an attempt to mimic the way a speaker reacts when a music signal is played through it.
Here's an old paper on Dynamic Convolution:
http://www.sintefex.com/docs/appnotes/dynaconv.PDF The paper is old, but the implementation of the idea is just beginning to happen on desktop DAW because average CPU power is now capable of performing the calculations effectively.
best regards,
mike