Decoupling of speakers is a challenge.
We all seem to worry about the phase accuracy of our speakers, but then we want to introduce a decoupling solution that by definition needs to incorporate some spring to be able to function as a decoupler.
I've long pondered this as, besides being in tech day-to-day, I have a side hustle that provides pretty serious vibration and acoustic isolation solutions to industry, studios, gyms, etc.
I've worked with a polyurethane material called Sylomer since about 2007, when I was building my own studio which went up in smoke thanks to 2008 financial crash.
When I'm speccing a solution for a client, we focus on the frequencies of excitation, what the target natural frequency for the system is as a whole, and the load applied by the machine itself.
Relatively simple when you're dealing with a generator that runs at 3,000rpm (50Hz), except for the startup and shutdown phases where there is obviously a rise up to operating rpm, and then a slowdown.
When dealing with something like a speaker though, I do worry about any micro-movements that may affect alignment. (Never tested this myself though)
It is the reason why the Mesanovic solution has never sat well with me.
Steel springs ring at higher frequencies when there is enough excitation, even though they can isolate very low down.
That ringing causes vibration to transmit through the rest of the isolation.
Northward Acoustics/Systems solution however is far more appealing to me.
The constrained mounting they have in their stands uses Sylomer, which provides both spring (for isolation) and damping (energy absorption).
(link for the interested:
northward-systems.com/product/speaker-stand-type-1525-kii-three/)
You can see in the specs that there is a natural frequency when minimally loaded (12Hz) and maximally loaded (8Hz).
That natural frequency point has a Q.
The Q for Sylomer is wider than the Q for steel springs, but it drops off relatively quickly.
Here is an example of a calculation for an ATC SCM45A Pro speaker:
The above calculation was done assuming:
1 x 36kg Speaker
4 x Isolation Pads measuring 50x50x50mm
The resulting natural frequency of the system as a whole is 9.1Hz, and at that frequency you can see an amplification of energy in the order of 500%.
But by the time you get up to 16Hz, there is a 51% attenuation.
At 20Hz, a 73% attenuation, etc.
(you can see this detail in the table).
The goal with decoupling is to stop the energy from the speaker from causing audible vibrations in other objects. (desks, stands, plasterboard walls and ceilings, etc.)
And although lower is generally better (steel springs will get a natural frequency below 5Hz without much fuss, but will transmit high frequencies due to its inherent stiffness), one really need only aim for getting a natural frequency enough below 20Hz to ensure that proper decoupling is done.
Sorry for the long post.
It is a passion of mine.
It's crazy to me that that works, always messes with my brain. Anyway, give it a try and report back! 
