There's an old medical adage that goes something like "When you hear hoof beats look for horses, not zebras."  Zebras are rare and exotic, while horses are much more common. It is often times best to think of common solutions first like a horse than to look for an exotic solution like a zebra.  This same adage applies to acceleration, start with the simple things first before attempting the more advanced options.   By starting with the common items you can start to see immediate benefits, the more exotic options require more time and resources as these often require more in-depth testing.  When deploying an acceleration solution the simple options include compression and static caching in an asymmetric deployment these will likely provide a good first pass and will provide value.  Slowly the more exotic options can be deployed such as dynamic caching, invalidations, and symmetric deployments.