The X post is obviously biased, but I couldn't spot any obvious scientific mistakes in the actual announcement video.
Someone else linked to this preprint which seems related [1]. Would you take a look and say whether it seems legitimate?
[1] Whole Cross-Sectional Human Ultrasound Tomography” https://arxiv.org/pdf/2307.00110
The authors report explicitly state that the system's resolution does not match clinical CT or MRI. The elevational resolution is 15–25mm, meaning each slice is effectively a thick 2D section rather than the fine isotropic 3D volume MRI provides. MRI also delivers far richer soft-tissue contrast; this device produces only three contrast types (reflectivity, speed of sound, attenuation), and because it uses a low 1MHz frequency, the reflection images come primarily from tissue boundaries rather than from internal tissue texture.
I could see this being valuable for adipose tissue mapping or fatty-liver monitoring at a large scale, as the machines would be significantly cheaper, but this isn't some revolutionary magic bullet like the Twitter post is insinuating.