Hi, author here -- I agree, the BladeRF is overkill for this kind of application. But, we were done with what we needed it for and it was just sitting around our lab so we decided to use it. I did use an RTL-SDR to poke around in the lower half of UHF, though.
edit: After inspecting the code, spectrum_sense is indeed applying a blackman-harris window before the DFT.