The whole vibration thing is interesting. Suppose you put a mass on something that is springy, and the other end is coupled to something rigid, like a floor. What is the resonance frequency? It is dead easy to calculate - if the spring compresses by delta x metres, the resonance frequency is (1/2pi) x root(g/delta x). You do not need to know the mass and spring constant - just the deflection. So suppose the speaker compresses (say) sorbothane feet by 2mm, the resonant frequency is (1/2pi) x root(9.81/2E-3) = 11Hz. So at frequencies higher than 11Hz the speaker stays (vertically) stationary, and at frequencies lower than 11Hz, the speaker bounces up and down. Of course if it is actually sorbothane, there is high loss, so there is not a resonant peak, but the basic motion remains.
Of course it is more complex than that, because the cones have substantial mass, so there will be a tendency for the speakers to rock like a pendulum as well as the cones move back and forth. In this case it depends on where the C of G of the speakers lie, where the drivers are (particularly the bass driver), and how far apart the springy bits (sorbothane for example) are. That in general will be a low frequency, certainly lower than the vertical 11Hz. Try rocking the speaker and see what happens - probably 1 to 2Hz.
If you use spikes there is a gotcha. The spikes need to be very, very seriously locked in place. Because of the cone motion, particularly at low frequency, and tendency to pendulum rock, it the spikes are not locked absolutely solid, the speaker can rock on the spikes. Most easily diagnosed with swept sine. Got the badge on that one - testing a homebrew sub I wondered what the god awful noises were as I swept a sine wave - sounded like the worst sort of distortion. Loose spikes.