In view of the ${\omega }^6$ scaling of the emitted power [Eq. (1)], going to much higher frequencies could help, but imagining masses $\mathit{m}=\mathcal{O}(\mathrm{kg})$ oscillating with amplitudes $\mathit{L}=\mathcal{O}(\mathrm{m})$ at frequencies in the kilohertz regime (and detecting small energy shifts of $\hslash \omega$) shows that such an experiment is also extremely challenging.

Charged particles such as protons and electrons can also be accelerated to nearly the speed of light, but in this case the electromagnetic radiation would dominate any gravitational effects by far.

Note that this applies to the standard coupling $h^{\mu \nu }{T}_{\mu \nu }$ to the energy-momentum tensor $T_{\mu \nu }$ in linearized general relativity. For other (hypothetical) coupling terms between photons and gravitons, the situation might be different. However, even in those cases, symmetries such as Lorentz invariance restrict such photon-graviton conversion processes strongly.

Even though the renormalized expectation values $⟨({\partial }_x{\widehat{A}}_z{)}^2{⟩}_{\mathrm{ren}}$ or $⟨({\partial }_y{\widehat{A}}_z{)}^2{⟩}_{\mathrm{ren}}$ could also become negative in some regions of space-time, the total volume integral of them must be non-negative.

These mirrors, however, change the propagation direction not from ${\mathit{K}}_{\mathrm{in}}=K_x{\mathit{e}}_x$ to ${\mathit{K}}_{\mathrm{out}}=K_y{\mathit{e}}_y$ but rather from ${\mathit{K}}_{\mathrm{in}}=K_x{\mathit{e}}_x$ to ${\mathit{K}}_{\mathrm{out}}=-K_x{\mathit{e}}_x$ or from ${\mathit{K}}_{\mathrm{in}}=K_y{\mathit{e}}_y$ to ${\mathit{K}}_{\mathrm{out}}=-K_y{\mathit{e}}_y$, and thus they do not change the sign of $\dot{E}$ in Eq. (5).

In principle, the stimulated emission or absorption of gravitons could also be achieved via mechanical means, e.g., via a rotating and vibrating barbell; see [16]. However, somewhat similar to a comment in [8], the required parameters make such an experiment quite challenging.