There really is no such thing as "vibration elimination" in an audio system.
There are various methods of trying to control the unwanted vibrations, or reduce the audibly harmful effects of the vibrations.
Nearly every little part in an audio system vibrates in some way or other, as part of its normal operation, even if the vibrations are very very small. These are things that comprise the equipment itself, and then we have the other vibrations that are a result of playing the equipment in the room.
Obviously, we don't want to eliminate all vibrations, or the system would not operate. Just imagine if all vibrations were "eliminated" from your speaker radiators. There would be no sound. So, what we are trying to do is remove the negative sounding effects of unwanted vibrations affecting our playback system.
There are various methods used by audiophiles to do this job, and they have various levels of effectiveness.
Both typically accepted methods - damping is one, and transfer is the other, have certain results of achieving the desired goal. Neither is perfect.
If you decide to use damping, then the method of damping and the amount of damping material used must be capable of performing the task at hand. The material must be able to absorb and internally damp the vibrational energy without reflecting it back into the component, and it must be placed in areas of best effect.
If you decide to use energy transfer(which provides a path for vibrational energy to exit the components, to be damped at another location, like the ground) then the items used to create this exit path must be capable of properly coupling to the components and able to provide a rapid enough route for them to follow to the destination, without causing the vibrations to be reflected back into the components.
Both of these methods require that a well-engineered and executed application be used for best results. While any method(such as tennis balls or rubber feet) may/will have some effect, it is totally unpredictable how they will actually perform in any given application.
If you use an engineered product which is known to work from historical and scientific background, and has models which can be shown to work in a similar application to the one you want, then you have the best chance of success. If you DIY from ad-hoc materials at hand, then you'll need to do some study to find out how much these things can handle, and what their expected results should be. Or you can just "play around" with stuff to see what you like.
In many cases, it actually winds up being cheaper and better to use something that a company has spent the time, research, testing, and consumer acceptance, in the first place to find out it works, than it might be to buy tons of stuff experimenting with it. But some find it to be a fun pastime to experiment with things.
Whichever method you choose, try to be careful to employ the items in a well-thought out manner, so that the technology that you use at least have a chance to give you good results. For example, putting an Audiopoint on top of a rubber block is not a good idea, if you want the Audiopoint to do what it is designed to do.
Whether you prefer the sound of one method or other, depends alot on your listening tastes, and the sounds of the things in your system. Not all things that can be used for either of these methods are created equal, either.
In point of fact, both the "damping" and "transfer" methods actually employ both technologies in one form or other, to work as intended. In any "damping" system, there is by necessity, some transfer of the vibrations to the areas which have the damping materials. This may be a short section of the component chassis, or whatever. So, there actually is some "transfer" in all damping systems. In "transfer" systems, the energy is transferred rapidly to another location, where it is naturally "damped" by the large masses of the house/building/ground. So, there is "damping" occurring in all "transfer" systems too. The main factors here, are where the damping is occurring, how much damping occurs locally or elsewhere, whether overload is occurring (because of too-slow of a tranfer ability in the case of transfer systems, or not enough damping ability in the case of damping systems) and how these results affect the purity of the musical presentation. Since local damping can affect the listening vibrations that we want to hear, we must also be careful to not "over-damp", and thus affect the musical presentation.
Disclaimer: I work for Starsound Technologies, maker of Audiopoints and Sistrum Platforms, and our company stands firmly in the Resonance Energy Transfer camp of ideologies regarding vibrational control in audio/video systems.