I'm highly skeptical that this hasn't already been done, considering that acousto-optic lenses are a known thing -- it's just a question of which material you're using (air, crystal, liquid, etc): https://en.wikipedia.org/wiki/Acousto-optic_modulator
Even a 10-second Google search yielded examples of prior art. For example:
„Modern photonics […] can involve parameter regimes where the wavelength or high optical powers involved restrict control due to absorption, light-induced damage or optical nonlinearity in solid media. Here we propose to circumvent these constraints using gaseous media tailored by high-intensity ultrasound waves. We demonstrate an implementation of this approach by efficiently deflecting ultrashort laser pulses using ultrasound waves in ambient air, without the use of transmissive solid media. At optical peak powers of 20 GW, exceeding previous limits of solid-based acousto-optic modulation by about three orders of magnitude, we reach a deflection efficiency greater than 50% while preserving excellent beam quality.“
I wondered the same thing. Acousto-optic modulators and deflectors have been commercial items since the late 1970s/early 1980s. I used them in that era. They were even intrinsic parts of laser systems, sitting within the cavity to create pulses of various kinds.
There's also a 50 year history of measuring dynamics in gas and condensed phase systems by the transient grating method which uses laser pulses to generate temporary gratings in all sorts of media. Two short (nanoseconds and on down) cross in a sample (solid, liquid, gas, flame, etc.). They generate an interference pattern and the fringe separation depends on the crossing angle. At points of high intensity, some kind of excitation is created and the refractive index gets modulated by the fringe pattern forming a grating. Now bring in a third pulse at a later time and it will scatter off of this grating. The grating will decay with time as the excitation relaxes due to whatever physical or chemical process is under study. Change the delay between the excitation and probe pulses to get a decay curve which is the observable to be fitted, modeled, compared with theory, etc.
It's different because they are using a gas as an acousto-optic medium. This allowed them to work with laser intensities that are more than three orders of magnitude above the damage threshold of conventional solid-state acousto-optic modulators.
Achieving good diffraction efficiency (they report ~50%) with a gaseous medium is a serious engineering challenge since the refractive index variations due to sound waves are many orders of magnitude weaker than in solids.
I would suggest reading the original publication [1] for more details. It's open access.
Even a 10-second Google search yielded examples of prior art. For example:
https://iopscience.iop.org/article/10.1088/2515-7647/abc23c/...
How is this any different? ¯\_(ツ)_/¯