Imagine you're sitting at a piano, pressing a key and listening for the hammer to strike the string. If the linkage were perfectly rigid, the sound would arrive the instant you pressed. But if there's even a tiny bit of slop — a spring, a damper — you'd notice a lag that grows worse the faster you try to play. This paper proposes doing exactly that test with gravity: spin a mass fast enough that any "slop" in the gravitational linkage — any delay, any relaxation time — becomes measurable. The claim is that two magnetically levitated NdFeB rotors, one acting as a gravitational transmitter (spinning mass quadrupole) and the other as a detector (torque sensor), can probe whether the near-field gravitational coupling has a finite relaxation time, pushing constraints from the current 10⁻⁵ seconds down to 10⁻⁸ seconds. That's a three-order-of-magnitude improvement over the best existing bound, set at just 1.7 kHz. The key trick is elegant: the detector rotor acts as a mechanical local oscillator. When the transmitter spins at megahertz rates, the gravitational tidal torque it produces on the detector is modulated at twice the transmitter's rotation frequency. But because the detector is also spinning, the signal gets heterodyned — downconverted — to a low beat frequency of just a few hertz, which is far easier to measure. This is the same principle as a radio superheterodyne receiver, except the carrier is gravitational torque instead of electromagnetic waves. Crucially, the authors show that at fixed geometry, neither the gravitational torque amplitude nor the required integration time depends on the rotation rate. You spin faster, you probe higher frequencies, and it costs you nothing in signal-to-noise. The numbers are specific. At 2 MHz rotation, the predicted torque is 2×10⁻²⁵ Nm. At 10 mK temperature with realistic NdFeB rotor parameters (density ~7,400 kg/m³, radius ~100 μm), thermal noise is low enough that this torque is detectable in under an hour. The rotor masses and geometries are within reach of current magnetic levitation technology — no exotic hardware is required, though the cryogenic environment and magnetic shielding are nontrivial engineering challenges. The theoretical framework models any departure from instantaneous Newtonian coupling through a single phenomenological parameter: a gravitational field relaxation time τ. If gravity propagates truly instantaneously in the near field (as GR predicts for the static/quasistatic regime), the torque is frequency-independent and τ is consistent with zero. If there's a finite τ, the torque rolls off at frequencies above 1/τ, producing a measurable signature. The current best constraint is τ < 10⁻⁵ s from the 1.7 kHz measurement; this proposal would push to τ < 10⁻⁸ s. As a bonus, the authors derive the sensitivity of a single rotor operated as a tunable high-frequency gravitational-wave detector. The noise-equivalent strain they estimate is ~10⁻²² Hz⁻¹/², which is not competitive with LIGO at its frequencies but opens a detection window in the megahertz band where no detector currently operates. This is a secondary result but an interesting one — it positions the same hardware for a completely different physics program. The integrity picture is honest but limited: this is a proposal paper, not a measurement. There is no experimental data. The analysis rests on analytical calculations of Newtonian gravitational torques and standard thermal noise models. The authors are upfront about what they're assuming (rigid rotors, perfect alignment, thermal-noise-limited operation) and what they're not modeling (rotor imperfections, magnetic crosstalk, seismic isolation). The validation is self-consistent theoretical analysis, not independent experiment. The practical path forward is clear: someone needs to build it. The components exist individually — magnetically levitated NdFeB rotors have been demonstrated at high spin rates, cryogenic torque measurements are routine in optomechanics labs — but the specific configuration (two rotors, controlled geometry, MHz spin rates, 10 mK) has not been assembled. The gap between proposal and first measurement is probably 3-5 years of dedicated experimental effort, assuming a group with the right cryogenic and levitation infrastructure picks this up.
