What we've got here is a lot of marketing fluff around a Rydberg spectrum analyzer.[1][1a]. This is essentially an RF mixer[2] which, according to [1a]:

  rivals the performance of commercially-available spectrum analyzers with high sensitivity, 4 MHz instantaneous bandwidth, and over 80 dB of linear dynamic range.

While 80 dB of linear dynamic range is good, there are other designs, such as the Tayloe mixer[3][4] that are more sensitive and have a better dynamic range due to their inherently linear nature.

Careful reading of the linked paper shows that their best sensitivity measurements rely on conventional external low noise amplifiers. So, this is mostly grant writing driven marketing.

We've been getting signal detection this good with tube based WW2 and later ham gear my friend repairs. It's not uncommon to detect 0.1 μV signals, which is -127 dBm. These receivers have approximately 2khz of bandwidth

     -127 dBm - 10*Log(2000) --> -160dBm/Hz.

So, lots of interesting tangents, but nothing world changing as of now.

[1] https://arxiv.org/abs/2009.14383

[1a PDF] https://arxiv.org/pdf/2009.14383.pdf

[2] https://en.wikipedia.org/wiki/Frequency_mixer

[3 pdf] https://www.norcalqrp.org/files/Tayloe_mixer_x3a.pdf

[4] https://pa3fwm.nl/technotes/tn17c.html