24 GHz and 60 GHz are significantly more expensive to deal with – whilst your average FR4 PCB is totally sufficient in material \$\epsilon\$ uniformity and \$\tan\delta\$, it won't do > 20 GHz.
Reliable connectors become very expensive.
Also, you either get much more loss per distance, or you use the reduced wavelength to build an antenna with larger directivity that's the same size than its 2.4 GHz counterpart. That makes sense for radar, but for communications, you don't want that (unless you can build even more complicated digital beamforming systems – a questions of cost and complexity).
New wireless broadband communication systems going to 60 GHz is a deliberate choice to go for the quasioptical channel, that solves a couple of the issues of the less selective / specular 2.4 GHz indoor channel out of necessity. You simply can't do future bandwidths in 2.4 GHz. If we could, we would. By going higher, we get drastically increased free-space loss and much narrower fading "holes", but we get the ability to do massive MIMO in indoors scenarios and get Gigahertzes of bandwidth. 24 GHz has a meagre 250 MHz of bandwidth – not that much more than 5 GHz already offers.
24 GHz is simply a bad tradeoff between much more cost and complexity vs advantages; you don't get the quasi-optical diversity, but you get all the mmWave complications.